<?xml version="1.0"?>
<feed xmlns="http://www.w3.org/2005/Atom" xml:lang="pt">
	<id>https://aprendis.med.up.pt/index.php?action=history&amp;feed=atom&amp;title=Text_Mining</id>
	<title>Text Mining - Histórico de revisões</title>
	<link rel="self" type="application/atom+xml" href="https://aprendis.med.up.pt/index.php?action=history&amp;feed=atom&amp;title=Text_Mining"/>
	<link rel="alternate" type="text/html" href="https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;action=history"/>
	<updated>2026-08-14T06:11:00Z</updated>
	<subtitle>Histórico de edições para esta página nesta wiki</subtitle>
	<generator>MediaWiki 1.39.1</generator>
	<entry>
		<id>https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3578&amp;oldid=prev</id>
		<title>Franzbischoff em 14h57min de 23 de março de 2016</title>
		<link rel="alternate" type="text/html" href="https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3578&amp;oldid=prev"/>
		<updated>2016-03-23T14:57:34Z</updated>

		<summary type="html">&lt;p&gt;&lt;/p&gt;
&lt;table style=&quot;background-color: #fff; color: #202122;&quot; data-mw=&quot;interface&quot;&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;tr class=&quot;diff-title&quot; lang=&quot;pt&quot;&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;← Revisão anterior&lt;/td&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;Revisão das 15h57min de 23 de março de 2016&lt;/td&gt;
				&lt;/tr&gt;&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l25&quot;&gt;Linha 25:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Linha 25:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[Ficheiro:Bischoff text mining framework.png|thumb|right|Figura 1: Framework do text mining (Adaptado de Tan ''et al.''&amp;lt;ref name=&amp;quot;Tan1999&amp;quot;/&amp;gt;).]]&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[Ficheiro:Bischoff text mining framework.png|thumb|right|Figura 1: Framework do text mining (Adaptado de Tan ''et al.''&amp;lt;ref name=&amp;quot;Tan1999&amp;quot;/&amp;gt;).]]&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;As decisões sobre a forma de representar os documentos podem ter um impacto significativo na qualidade dos padrões posteriormente identificados. Existem alguns métodos que são mais &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;comumente &lt;/del&gt;utilizados para representar o texto:&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;As decisões sobre a forma de representar os documentos podem ter um impacto significativo na qualidade dos padrões posteriormente identificados. Existem alguns métodos que são mais &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;comummente &lt;/ins&gt;utilizados para representar o texto:&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Elementos ao nível de superfície:''' captura informações sobre uma palavra. Por exemplo, nomes próprios normalmente começam com uma letra maiúscula;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Elementos ao nível de superfície:''' captura informações sobre uma palavra. Por exemplo, nomes próprios normalmente começam com uma letra maiúscula;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Representação baseada em vetor:''' a representação de texto mais &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;comumente &lt;/del&gt;utilizada em texto mining é a abordagem de saco de palavras (''bag of words'' - BOW). A representação BOW também se encaixa bem em métodos de análise que requerem uma medida de similaridade entre documentos (como clustering);&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Representação baseada em vetor:''' a representação de texto mais &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;comummente &lt;/ins&gt;utilizada em texto mining é a abordagem de saco de palavras (''bag of words'' - BOW). A representação BOW também se encaixa bem em métodos de análise que requerem uma medida de similaridade entre documentos (como clustering);&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Representação de conceito:''' Uma boa representação de texto deve resolver sinonímias (quando diferentes palavras têm o mesmo significado) e polissemia (quando a mesma palavra tem significados diferentes). A Biblioteca Nacional de Medicina dos Estados Unidos da América (NLM) desenvolveu o Sistema de Linguagem Médica Única (UMLS). Similarmente, um conceito pode ser rastreado até o vocabulário de origem e os termos são mapeados para um ou mais conceitos a níveis mais elevados chamados de tipos semânticos&amp;lt;ref name=&amp;quot;Aronson2001&amp;quot;&amp;gt;Aronson a R: Effective mapping of biomedical text to the UMLS Metathesaurus: the MetaMap program. Proc AMIA Symp 2001:17–21.&amp;lt;/ref&amp;gt;, que podem ser utilizados para resolver a polissemia.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Representação de conceito:''' Uma boa representação de texto deve resolver sinonímias (quando diferentes palavras têm o mesmo significado) e polissemia (quando a mesma palavra tem significados diferentes). A Biblioteca Nacional de Medicina dos Estados Unidos da América (NLM) desenvolveu o Sistema de Linguagem Médica Única (UMLS). Similarmente, um conceito pode ser rastreado até o vocabulário de origem e os termos são mapeados para um ou mais conceitos a níveis mais elevados chamados de tipos semânticos&amp;lt;ref name=&amp;quot;Aronson2001&amp;quot;&amp;gt;Aronson a R: Effective mapping of biomedical text to the UMLS Metathesaurus: the MetaMap program. Proc AMIA Symp 2001:17–21.&amp;lt;/ref&amp;gt;, que podem ser utilizados para resolver a polissemia.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l43&quot;&gt;Linha 43:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Linha 43:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Radicalização (ou lematização):''' a radicalização tenta identificar a raiz de um termo. Por exemplo, os termos &amp;quot;diferem&amp;quot;, &amp;quot;diferentes&amp;quot; e &amp;quot;diferença&amp;quot; seriam todos representados como &amp;quot;difer&amp;quot; após ser aplicada a radicalização.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Radicalização (ou lematização):''' a radicalização tenta identificar a raiz de um termo. Por exemplo, os termos &amp;quot;diferem&amp;quot;, &amp;quot;diferentes&amp;quot; e &amp;quot;diferença&amp;quot; seriam todos representados como &amp;quot;difer&amp;quot; após ser aplicada a radicalização.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Palavras frequentes (''stop words''):''' termos que ocorrem com &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;freqüência &lt;/del&gt;no texto, mas têm pouco poder discriminatório como determinantes, preposições, etc.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Palavras frequentes (''stop words''):''' termos que ocorrem com &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;frequência &lt;/ins&gt;no texto, mas têm pouco poder discriminatório como determinantes, preposições, etc.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Poda de frequência:''' a remoção de palavras que aparecem em menos de n documentos (onde N é geralmente em torno de 5) ou mais do que m vezes (onde m é geralmente uma percentagem do número de documentos).&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Poda de frequência:''' a remoção de palavras que aparecem em menos de n documentos (onde N é geralmente em torno de 5) ou mais do que m vezes (onde m é geralmente uma percentagem do número de documentos).&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Indexação semântica latente:''' é uma técnica estatística que, dada uma matriz de documentos e termos, gera uma série de vetores que capturam a variância dentro da matriz original&amp;lt;ref name=&amp;quot;Deerwester1990&amp;quot;&amp;gt;Deerwester S, Dumais ST, Furnas GW, Landauer TK, Harshman R: Indexing by latent semantic analysis. J Am Soc Inf Sci 1990, 41:391–407.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Indexação semântica latente:''' é uma técnica estatística que, dada uma matriz de documentos e termos, gera uma série de vetores que capturam a variância dentro da matriz original&amp;lt;ref name=&amp;quot;Deerwester1990&amp;quot;&amp;gt;Deerwester S, Dumais ST, Furnas GW, Landauer TK, Harshman R: Indexing by latent semantic analysis. J Am Soc Inf Sci 1990, 41:391–407.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l61&quot;&gt;Linha 61:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Linha 61:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Coesão textual (''Textual Entailment''):''' em NLP, a coesão textual é uma relação direcional entre dois fragmentos de texto. Um dos objetivos é identificar se uma dada hipótese (H) está relacionada com uma dada expressão (T). A hipótese geralmente tem a forma de uma sentença única, e o texto pode ser de uma ou mais sentenças.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Coesão textual (''Textual Entailment''):''' em NLP, a coesão textual é uma relação direcional entre dois fragmentos de texto. Um dos objetivos é identificar se uma dada hipótese (H) está relacionada com uma dada expressão (T). A hipótese geralmente tem a forma de uma sentença única, e o texto pode ser de uma ou mais sentenças.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Sumarização:''' desde o final dos anos sessenta, procuram-se construir sistemas que geram automaticamente um resumo de um único documento através da seleção de texto já existente ou abstraindo o texto dos documentos originais, transformando e gerando novo texto.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Sumarização:''' desde o final dos anos sessenta, procuram-se construir sistemas que geram automaticamente um resumo de um único documento através da seleção de texto já existente ou abstraindo o texto dos documentos originais, transformando e gerando novo texto.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Síntese de informação:''' embora a sumarização de &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;multiplos &lt;/del&gt;documentos seja inerentemente uma atividade de síntese, as pesquisas sobre esta questão tem sido em torno do processo de revisão sistemática. O principal objetivo de uma meta-análise é equilibrar evidências contraditórias e redundantes na literatura científica.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Síntese de informação:''' embora a sumarização de &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;múltiplos &lt;/ins&gt;documentos seja inerentemente uma atividade de síntese, as pesquisas sobre esta questão tem sido em torno do processo de revisão sistemática. O principal objetivo de uma meta-análise é equilibrar evidências contraditórias e redundantes na literatura científica.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Descoberta de conhecimento baseada em literatura:''' semelhantemente à síntese de informação, a sobrecarga de informação na literatura biomédica levou ao desenvolvimento da descoberta de conhecimento baseada em literatura, que é um processo de pesquisa &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;sistematica &lt;/del&gt;da literatura existente para identificar novos conhecimentos.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Descoberta de conhecimento baseada em literatura:''' semelhantemente à síntese de informação, a sobrecarga de informação na literatura biomédica levou ao desenvolvimento da descoberta de conhecimento baseada em literatura, que é um processo de pesquisa &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;sistemática &lt;/ins&gt;da literatura existente para identificar novos conhecimentos.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Text Mining na Saúde ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Text Mining na Saúde ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l68&quot;&gt;Linha 68:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Linha 68:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;O termo ''Clinical Data Mining'' consiste na aplicação de técnicas de [[Data Mining|data mining]] utilizando dados clínicos&amp;lt;ref name=&amp;quot;Iavindrasana2009&amp;quot;&amp;gt;Iavindrasana J, Cohen G, Depeursinge A, Müller H, Meyer R, Geissbuhler A: Clinical data mining: a review. Yearb Med Inform 2009(December 2015):121–33.&amp;lt;/ref&amp;gt;. No campo biomédico, a importância do text mining se reflete no aumento de mais de seis vezes no número de publicações sobre text mining nas ultimas décadas. As companhias farmacêuticas consideram o text mining como uma &amp;quot;necessidade básica&amp;quot;&amp;lt;ref name=&amp;quot;Kumar2007&amp;quot;&amp;gt;Kumar VD, Tipney HJ: Biomedical Literature Mining. Volume 1159. New York, NY: Springer New York; 2007. [Methods in Molecular Biology]&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;O termo ''Clinical Data Mining'' consiste na aplicação de técnicas de [[Data Mining|data mining]] utilizando dados clínicos&amp;lt;ref name=&amp;quot;Iavindrasana2009&amp;quot;&amp;gt;Iavindrasana J, Cohen G, Depeursinge A, Müller H, Meyer R, Geissbuhler A: Clinical data mining: a review. Yearb Med Inform 2009(December 2015):121–33.&amp;lt;/ref&amp;gt;. No campo biomédico, a importância do text mining se reflete no aumento de mais de seis vezes no número de publicações sobre text mining nas ultimas décadas. As companhias farmacêuticas consideram o text mining como uma &amp;quot;necessidade básica&amp;quot;&amp;lt;ref name=&amp;quot;Kumar2007&amp;quot;&amp;gt;Kumar VD, Tipney HJ: Biomedical Literature Mining. Volume 1159. New York, NY: Springer New York; 2007. [Methods in Molecular Biology]&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Novas tecnologias tem sido desenvolvidas no &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;âmbido &lt;/del&gt;do ''Clinical Data Mining'', como por exemplo um novo método de anotar dados clínicos não estruturados e transformá-los em uma matrix codificada e anonimizada de paciente-características utilizando terminologias médicas com uma das finalidades de detectar efeitos colaterais de drogas e efeitos colaterais de efeitos de associação de drogas&amp;lt;ref name=&amp;quot;LePendu2013&amp;quot;&amp;gt;LePendu P, Iyer S V, Bauer-Mehren A, Harpaz R, Mortensen JM, Podchiyska T, Ferris TA, Shah NH: Pharmacovigilance Using Clinical Notes. Clin Pharmacol Ther 2013, 93:547–55.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Novas tecnologias tem sido desenvolvidas no &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;âmbito &lt;/ins&gt;do ''Clinical Data Mining'', como por exemplo um novo método de anotar dados clínicos não estruturados e transformá-los em uma matrix codificada e anonimizada de paciente-características utilizando terminologias médicas com uma das finalidades de detectar efeitos colaterais de drogas e efeitos colaterais de efeitos de associação de drogas&amp;lt;ref name=&amp;quot;LePendu2013&amp;quot;&amp;gt;LePendu P, Iyer S V, Bauer-Mehren A, Harpaz R, Mortensen JM, Podchiyska T, Ferris TA, Shah NH: Pharmacovigilance Using Clinical Notes. Clin Pharmacol Ther 2013, 93:547–55.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Eis algumas das aplicações do text mining na literatura biomédica:&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Eis algumas das aplicações do text mining na literatura biomédica:&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* Interações medicamentosas&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* Interações medicamentosas&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* Análise de co-&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;ocorrencias &lt;/del&gt;e extração de informações biológicas&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* Análise de co-&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;ocorrências &lt;/ins&gt;e extração de informações biológicas&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* Procura de novidades biológicas ao desvendar informações &amp;quot;ocultas&amp;quot;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* Procura de novidades biológicas ao desvendar informações &amp;quot;ocultas&amp;quot;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* Predição da função de proteínas&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* Predição da função de proteínas&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;/table&gt;</summary>
		<author><name>Franzbischoff</name></author>
	</entry>
	<entry>
		<id>https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3577&amp;oldid=prev</id>
		<title>Franzbischoff: Saúde e Desafios (by SublimeText.Mediawiker)</title>
		<link rel="alternate" type="text/html" href="https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3577&amp;oldid=prev"/>
		<updated>2016-03-23T14:49:24Z</updated>

		<summary type="html">&lt;p&gt;Saúde e Desafios (by SublimeText.Mediawiker)&lt;/p&gt;
&lt;table style=&quot;background-color: #fff; color: #202122;&quot; data-mw=&quot;interface&quot;&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;tr class=&quot;diff-title&quot; lang=&quot;pt&quot;&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;← Revisão anterior&lt;/td&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;Revisão das 15h49min de 23 de março de 2016&lt;/td&gt;
				&lt;/tr&gt;&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l65&quot;&gt;Linha 65:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Linha 65:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Text Mining na Saúde ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Text Mining na Saúde ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;O termo ''Clinical Data Mining'' consiste na aplicação de técnicas de [[Data Mining|data mining]] utilizando dados clínicos&amp;lt;ref name=&quot;Iavindrasana2009&quot;&gt;Iavindrasana J, Cohen G, Depeursinge A, Müller H, Meyer R, Geissbuhler A: Clinical data mining: a review. Yearb Med Inform 2009(December 2015):121–33.&amp;lt;/ref&gt;. No campo biomédico, a importância do text mining se reflete no aumento de mais de seis vezes no número de publicações sobre text mining nas ultimas décadas. As companhias farmacêuticas consideram o text mining como uma &quot;necessidade básica&quot;&amp;lt;ref name=&quot;Kumar2007&quot;&gt;Kumar VD, Tipney HJ: Biomedical Literature Mining. Volume 1159. New York, NY: Springer New York; 2007. [Methods in Molecular Biology]&amp;lt;/ref&gt;.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Novas tecnologias tem sido desenvolvidas no âmbido do ''Clinical Data Mining'', como por exemplo um novo método de anotar dados clínicos não estruturados e transformá-los em uma matrix codificada e anonimizada de paciente-características utilizando terminologias médicas com uma das finalidades de detectar efeitos colaterais de drogas e efeitos colaterais de efeitos de associação de drogas&amp;lt;ref name=&quot;LePendu2013&quot;&gt;LePendu P, Iyer S V, Bauer-Mehren A, Harpaz R, Mortensen JM, Podchiyska T, Ferris TA, Shah NH: Pharmacovigilance Using Clinical Notes. Clin Pharmacol Ther 2013, 93:547–55.&amp;lt;/ref&gt;.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Eis algumas das aplicações do text mining na literatura biomédica:&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Interações medicamentosas&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Análise de co-ocorrencias e extração de informações biológicas&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Procura de novidades biológicas ao desvendar informações &quot;ocultas&quot;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Predição da função de proteínas&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Unidades moleculares funcionais para guiar o desenho de painéis de biomarcadores&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Investigar redes biológicas a partir de textos completos de artigos&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Prever futuras descobertas através da literatura científica atual&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Compreender associações de doenças na descoberta de novas drogas&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Eleger genes candidados a certas doenças&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Identificação precoce de potenciais problemas de segurança farmacológica&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Conhecimento sobre doenças&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;== Desafios ==&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;As questões relacionadas ao text mining que foram expostas acima são problemas de alto nível. Há entretanto alguns desafios que devem ser solucionados em nível de aplicação e métodos individuais. Assim como problemas de semântica, as variações biológicas podem ser um fator complicante. Por exemplo, o que acontece com uma espécie não necessariamente acontece com outra, e mais ainda, pode não acontecer na mesma espécie quando em diferentes circunstâncias (género, genética, história natural, ou diferenças de terapêutica)&amp;lt;ref name=&quot;Kumar2007&quot;/&gt;.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Referências ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Referências ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;/table&gt;</summary>
		<author><name>Franzbischoff</name></author>
	</entry>
	<entry>
		<id>https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3576&amp;oldid=prev</id>
		<title>Franzbischoff: /* Refinação do Texto */</title>
		<link rel="alternate" type="text/html" href="https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3576&amp;oldid=prev"/>
		<updated>2016-03-23T14:19:40Z</updated>

		<summary type="html">&lt;p&gt;&lt;span dir=&quot;auto&quot;&gt;&lt;span class=&quot;autocomment&quot;&gt;Refinação do Texto&lt;/span&gt;&lt;/span&gt;&lt;/p&gt;
&lt;table style=&quot;background-color: #fff; color: #202122;&quot; data-mw=&quot;interface&quot;&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;tr class=&quot;diff-title&quot; lang=&quot;pt&quot;&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;← Revisão anterior&lt;/td&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;Revisão das 15h19min de 23 de março de 2016&lt;/td&gt;
				&lt;/tr&gt;&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l21&quot;&gt;Linha 21:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Linha 21:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;=== Refinação do Texto ===&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;=== Refinação do Texto ===&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Consiste em transformar texto livre em um formato intermediário, que pode ser baseado em documento ou baseado em conceitos. Na Figura 1 temos uma visão global do processo&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;:&lt;/del&gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Consiste em transformar texto livre em um formato intermediário, que pode ser baseado em documento ou baseado em conceitos. Na Figura 1 temos uma visão global do processo&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;FIGURA &lt;/del&gt;1: Framework do text mining (Adaptado de Tan ''et al.''&amp;lt;ref name=&amp;quot;Tan1999&amp;quot;/&amp;gt;).&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;[[Ficheiro:Bischoff text mining framework.png|thumb|right|Figura &lt;/ins&gt;1: Framework do text mining (Adaptado de Tan ''et al.''&amp;lt;ref name=&amp;quot;Tan1999&amp;quot;/&amp;gt;).&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;]]&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;As decisões sobre a forma de representar os documentos podem ter um impacto significativo na qualidade dos padrões posteriormente identificados. Existem alguns métodos que são mais comumente utilizados para representar o texto:&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;As decisões sobre a forma de representar os documentos podem ter um impacto significativo na qualidade dos padrões posteriormente identificados. Existem alguns métodos que são mais comumente utilizados para representar o texto:&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;/table&gt;</summary>
		<author><name>Franzbischoff</name></author>
	</entry>
	<entry>
		<id>https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3574&amp;oldid=prev</id>
		<title>Franzbischoff: Destilação do Conhecimento (by SublimeText.Mediawiker)</title>
		<link rel="alternate" type="text/html" href="https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3574&amp;oldid=prev"/>
		<updated>2016-03-23T12:31:29Z</updated>

		<summary type="html">&lt;p&gt;Destilação do Conhecimento (by SublimeText.Mediawiker)&lt;/p&gt;
&lt;table style=&quot;background-color: #fff; color: #202122;&quot; data-mw=&quot;interface&quot;&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;tr class=&quot;diff-title&quot; lang=&quot;pt&quot;&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;← Revisão anterior&lt;/td&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;Revisão das 13h31min de 23 de março de 2016&lt;/td&gt;
				&lt;/tr&gt;&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l48&quot;&gt;Linha 48:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Linha 48:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;=== Destilação do Conhecimento ===&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;=== Destilação do Conhecimento ===&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Após efetuada a representação de texto e a redução de dimensionalidade, o pesquisador analisa a coleção de textos. Abaixo verifica-se alguns métodos de análise independente de domínio que têm sido desenvolvidos para conjuntos de dados não-textuais:&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* '''Classificação (ou categorização):''' um algoritmo de classificação visa criar um modelo (o classificador) que mapeia com precisão, a partir de um conjunto de elementos usados para representar cada documento, para uma classe existente.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* '''Agrupamento (''clustering''):''' o objetivo de um algoritmo de ''clustering'' é agrupar documentos de tal forma que cada grupo tenha um alto grau de similaridade intra-classe e baixo grau de similaridade inter-classe. Em contraste com a classificação, as classes são desconhecidas.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Regras de associação: as regras de associação capturam co-ocorrências entre termos de uma coleção de documentos. Tais regras são muitas vezes geradas usando o algoritmo Apriori&amp;lt;ref name=&quot;Agrawal1996&quot;&gt;Agrawal R, Mannila H, Srikant R, Toivonen H, Verkamo a I: Fast discovery of association rules. Advances in knowledge discovery and data mining 1996:307–328.&amp;lt;/ref&gt;, que apresenta padrões para o usuário na forma: A?B, onde A e B são conjuntos de termos.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Todos os métodos anteriores de transformação de dados e análise deve ter um propósito. A seguir exemplifica-se alguns dos problemas mais comuns:&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* '''Extração de relação:''' o objetivo é identificar as relações entre os dados recolhidos. Por exemplo, as relações gene-proteína, detecção de hipônimos, relações de conceito como &quot;consiste em&quot;, &quot;feito de&quot;, etc.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* '''Detectar subjetividade:'''em linguagem natural define-se subjetividade como &quot;aspectos da linguagem utilizados para expressar opiniões, avaliações e especulações&quot;. Detectar subjetividade pode ser útil para classificações automáticas em diversos contextos como política, avaliação de produtos, opinião pública, etc.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* '''Coesão textual (''Textual Entailment''):''' em NLP, a coesão textual é uma relação direcional entre dois fragmentos de texto. Um dos objetivos é identificar se uma dada hipótese (H) está relacionada com uma dada expressão (T). A hipótese geralmente tem a forma de uma sentença única, e o texto pode ser de uma ou mais sentenças.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* '''Sumarização:''' desde o final dos anos sessenta, procuram-se construir sistemas que geram automaticamente um resumo de um único documento através da seleção de texto já existente ou abstraindo o texto dos documentos originais, transformando e gerando novo texto.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* '''Síntese de informação:''' embora a sumarização de multiplos documentos seja inerentemente uma atividade de síntese, as pesquisas sobre esta questão tem sido em torno do processo de revisão sistemática. O principal objetivo de uma meta-análise é equilibrar evidências contraditórias e redundantes na literatura científica.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* '''Descoberta de conhecimento baseada em literatura:''' semelhantemente à síntese de informação, a sobrecarga de informação na literatura biomédica levou ao desenvolvimento da descoberta de conhecimento baseada em literatura, que é um processo de pesquisa sistematica da literatura existente para identificar novos conhecimentos.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Text Mining na Saúde ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Text Mining na Saúde ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;/table&gt;</summary>
		<author><name>Franzbischoff</name></author>
	</entry>
	<entry>
		<id>https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3573&amp;oldid=prev</id>
		<title>Franzbischoff em 10h51min de 23 de março de 2016</title>
		<link rel="alternate" type="text/html" href="https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3573&amp;oldid=prev"/>
		<updated>2016-03-23T10:51:07Z</updated>

		<summary type="html">&lt;p&gt;&lt;/p&gt;
&lt;table style=&quot;background-color: #fff; color: #202122;&quot; data-mw=&quot;interface&quot;&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;tr class=&quot;diff-title&quot; lang=&quot;pt&quot;&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;← Revisão anterior&lt;/td&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;Revisão das 11h51min de 23 de março de 2016&lt;/td&gt;
				&lt;/tr&gt;&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l45&quot;&gt;Linha 45:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Linha 45:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Palavras frequentes (''stop words''):''' termos que ocorrem com freqüência no texto, mas têm pouco poder discriminatório como determinantes, preposições, etc.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Palavras frequentes (''stop words''):''' termos que ocorrem com freqüência no texto, mas têm pouco poder discriminatório como determinantes, preposições, etc.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Poda de frequência:''' a remoção de palavras que aparecem em menos de n documentos (onde N é geralmente em torno de 5) ou mais do que m vezes (onde m é geralmente uma percentagem do número de documentos).&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* '''Poda de frequência:''' a remoção de palavras que aparecem em menos de n documentos (onde N é geralmente em torno de 5) ou mais do que m vezes (onde m é geralmente uma percentagem do número de documentos).&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* Indexação semântica latente: é uma técnica estatística que, dada uma matriz de documentos e termos, gera uma série de vetores que capturam a variância dentro da matriz original&amp;lt;ref name=&amp;quot;Deerwester1990&amp;quot;&amp;gt;Deerwester S, Dumais ST, Furnas GW, Landauer TK, Harshman R: Indexing by latent semantic analysis. J Am Soc Inf Sci 1990, 41:391–407.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;'''&lt;/ins&gt;Indexação semântica latente:&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;''' &lt;/ins&gt;é uma técnica estatística que, dada uma matriz de documentos e termos, gera uma série de vetores que capturam a variância dentro da matriz original&amp;lt;ref name=&amp;quot;Deerwester1990&amp;quot;&amp;gt;Deerwester S, Dumais ST, Furnas GW, Landauer TK, Harshman R: Indexing by latent semantic analysis. J Am Soc Inf Sci 1990, 41:391–407.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;=== Destilação do Conhecimento ===&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;=== Destilação do Conhecimento ===&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;/table&gt;</summary>
		<author><name>Franzbischoff</name></author>
	</entry>
	<entry>
		<id>https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3572&amp;oldid=prev</id>
		<title>Franzbischoff: Refinação do Texto (by SublimeText.Mediawiker)</title>
		<link rel="alternate" type="text/html" href="https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3572&amp;oldid=prev"/>
		<updated>2016-03-23T10:50:32Z</updated>

		<summary type="html">&lt;p&gt;Refinação do Texto (by SublimeText.Mediawiker)&lt;/p&gt;
&lt;table style=&quot;background-color: #fff; color: #202122;&quot; data-mw=&quot;interface&quot;&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;tr class=&quot;diff-title&quot; lang=&quot;pt&quot;&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;← Revisão anterior&lt;/td&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;Revisão das 11h50min de 23 de março de 2016&lt;/td&gt;
				&lt;/tr&gt;&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l27&quot;&gt;Linha 27:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Linha 27:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;As decisões sobre a forma de representar os documentos podem ter um impacto significativo na qualidade dos padrões posteriormente identificados. Existem alguns métodos que são mais comumente utilizados para representar o texto:&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;As decisões sobre a forma de representar os documentos podem ter um impacto significativo na qualidade dos padrões posteriormente identificados. Existem alguns métodos que são mais comumente utilizados para representar o texto:&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* Elementos ao nível de superfície: captura informações sobre uma palavra. Por exemplo, nomes próprios normalmente começam com uma letra maiúscula;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;'''&lt;/ins&gt;Elementos ao nível de superfície:&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;''' &lt;/ins&gt;captura informações sobre uma palavra. Por exemplo, nomes próprios normalmente começam com uma letra maiúscula;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* Representação baseada em vetor: a representação de texto mais comumente utilizada em texto mining é a abordagem de saco de palavras (''bag of words'' - BOW). A representação BOW também se encaixa bem em métodos de análise que requerem uma medida de similaridade entre documentos (como clustering);&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;'''&lt;/ins&gt;Representação baseada em vetor:&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;''' &lt;/ins&gt;a representação de texto mais comumente utilizada em texto mining é a abordagem de saco de palavras (''bag of words'' - BOW). A representação BOW também se encaixa bem em métodos de análise que requerem uma medida de similaridade entre documentos (como clustering);&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* Representação de conceito: Uma boa representação de texto deve resolver sinonímias (quando diferentes palavras têm o mesmo significado) e polissemia (quando a mesma palavra tem significados diferentes). A Biblioteca Nacional de Medicina dos Estados Unidos da América (NLM) desenvolveu o Sistema de Linguagem Médica Única (UMLS). Similarmente, um conceito pode ser rastreado até o vocabulário de origem e os termos são mapeados para um ou mais conceitos a níveis mais elevados chamados de tipos semânticos&amp;lt;ref name=&amp;quot;Aronson2001&amp;quot;&amp;gt;Aronson a R: Effective mapping of biomedical text to the UMLS Metathesaurus: the MetaMap program. Proc AMIA Symp 2001:17–21.&amp;lt;/ref&amp;gt;, que podem ser utilizados para resolver a polissemia.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;* &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;'''&lt;/ins&gt;Representação de conceito:&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;''' &lt;/ins&gt;Uma boa representação de texto deve resolver sinonímias (quando diferentes palavras têm o mesmo significado) e polissemia (quando a mesma palavra tem significados diferentes). A Biblioteca Nacional de Medicina dos Estados Unidos da América (NLM) desenvolveu o Sistema de Linguagem Médica Única (UMLS). Similarmente, um conceito pode ser rastreado até o vocabulário de origem e os termos são mapeados para um ou mais conceitos a níveis mais elevados chamados de tipos semânticos&amp;lt;ref name=&amp;quot;Aronson2001&amp;quot;&amp;gt;Aronson a R: Effective mapping of biomedical text to the UMLS Metathesaurus: the MetaMap program. Proc AMIA Symp 2001:17–21.&amp;lt;/ref&amp;gt;, que podem ser utilizados para resolver a polissemia.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;De modo a representar um documento, a informação deve ser corretamente extraída a partir do documento. Várias estratégias têm sido utilizadas para realizar essa tarefa, e geralmente utiliza abordagens de aprendizagem de maquina (''machine learning''), baseado em conhecimento (''knowledge-based approach'') ou ambos (TEG&amp;lt;ref name=&quot;Feldman2006&quot;&gt;Feldman R, Rosenfeld B, Fresko M: TEG—a hybrid approach to information extraction. Knowl Inf Syst 2006, 9:1–18.&amp;lt;/ref&gt;).&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Extração de Informação (IE), é talvez atualmente a técnica mais utilizada em operações de pré-processamento em text mining. Sem as técnicas de IE, os sistemas de text mining teriam capacidades de descoberta de conhecimento muito mais limitadas. Mesmo os sistemas que não têm uma boa performance (''precision'' e ''recall'') podem ser de grande valor&amp;lt;ref name=&quot;Maletic2010&quot;/&gt;. A seguir lista-se os métodos mais utilizados:&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* '''Resolução de co-referência:''' o objetivo da resolução de co-referência é unificar todas as referências de texto para a mesma entidade (também chamada de referência), em que uma entidade é tipicamente um nome próprio, mas também podem ser relações verbais (ver Day ''et al.''&amp;lt;ref name=&quot;Hirschman1997&quot;&gt;Day D, Aberdeen J, Hirschman L, Kozierok R, Robinson P, Vilain M: Mixed-initiative development of language processing systems. In Proceedings of the fifth conference on Applied natural language processing -. Morristown, NJ, USA: Association for Computational Linguistics; 1997:348–355.&amp;lt;/ref&gt; para uma definição de anotações de co-referência utilizadas na tarefa de co-referência MUC-7).&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* '''Abreviaturas:''' as abreviaturas são um caso especial de resolução de co-referência e são particularmente importantes quando se trabalha com literatura científica. Um estudo descobriu que 80% das abreviaturas no UMLS eram ambíguas&amp;lt;ref name=&quot;Liu2002&quot;&gt;Liu H, Aronson AR, Friedman C: A study of abbreviations in MEDLINE abstracts. Proc AMIA Symp 2002:464–8.&amp;lt;/ref&gt;.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* '''Negação:''' a negação recebeu uma atenção especial nos registos médicos onde a pseudo negação, como &quot;nenhuma mudança&quot; e &quot;não causam&quot;, pode ter um enorme impacto sobre a qualidade dos padrões produzidos.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* '''Sintaxe:''' em contraste com as representações que incidem sobre a semântica, uma representação sintática captura as informações ao nível da sentença. Uma área onde a sintaxe pode ser útil é no resumo de documentos e síntese de informações.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Para se reduzir o ruído no texto original e, assim, proporcionar melhores padrões, pode-se utilizar técnicas de redução de dimensionalidade:&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* '''Radicalização (ou lematização):''' a radicalização tenta identificar a raiz de um termo. Por exemplo, os termos &quot;diferem&quot;, &quot;diferentes&quot; e &quot;diferença&quot; seriam todos representados como &quot;difer&quot; após ser aplicada a radicalização.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* '''Palavras frequentes (''stop words''):''' termos que ocorrem com freqüência no texto, mas têm pouco poder discriminatório como determinantes, preposições, etc.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* '''Poda de frequência:''' a remoção de palavras que aparecem em menos de n documentos (onde N é geralmente em torno de 5) ou mais do que m vezes (onde m é geralmente uma percentagem do número de documentos).&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Indexação semântica latente: é uma técnica estatística que, dada uma matriz de documentos e termos, gera uma série de vetores que capturam a variância dentro da matriz original&amp;lt;ref name=&quot;Deerwester1990&quot;&gt;Deerwester S, Dumais ST, Furnas GW, Landauer TK, Harshman R: Indexing by latent semantic analysis. J Am Soc Inf Sci 1990, 41:391–407.&amp;lt;/ref&gt;.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;=== Destilação do Conhecimento ===&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;=== Destilação do Conhecimento ===&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;/table&gt;</summary>
		<author><name>Franzbischoff</name></author>
	</entry>
	<entry>
		<id>https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3571&amp;oldid=prev</id>
		<title>Franzbischoff: WIP (by SublimeText.Mediawiker)</title>
		<link rel="alternate" type="text/html" href="https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3571&amp;oldid=prev"/>
		<updated>2016-03-23T10:15:04Z</updated>

		<summary type="html">&lt;p&gt;WIP (by SublimeText.Mediawiker)&lt;/p&gt;
&lt;table style=&quot;background-color: #fff; color: #202122;&quot; data-mw=&quot;interface&quot;&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;tr class=&quot;diff-title&quot; lang=&quot;pt&quot;&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;← Revisão anterior&lt;/td&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;Revisão das 11h15min de 23 de março de 2016&lt;/td&gt;
				&lt;/tr&gt;&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l20&quot;&gt;Linha 20:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Linha 20:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;=== Refinação do Texto ===&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;=== Refinação do Texto ===&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Consiste em transformar texto livre em um formato intermediário, que pode ser baseado em documento ou baseado em conceitos. Na Figura 1 temos uma visão global do processo:&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;FIGURA 1: Framework do text mining (Adaptado de Tan ''et al.''&amp;lt;ref name=&quot;Tan1999&quot;/&gt;).&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;As decisões sobre a forma de representar os documentos podem ter um impacto significativo na qualidade dos padrões posteriormente identificados. Existem alguns métodos que são mais comumente utilizados para representar o texto:&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Elementos ao nível de superfície: captura informações sobre uma palavra. Por exemplo, nomes próprios normalmente começam com uma letra maiúscula;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Representação baseada em vetor: a representação de texto mais comumente utilizada em texto mining é a abordagem de saco de palavras (''bag of words'' - BOW). A representação BOW também se encaixa bem em métodos de análise que requerem uma medida de similaridade entre documentos (como clustering);&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;* Representação de conceito: Uma boa representação de texto deve resolver sinonímias (quando diferentes palavras têm o mesmo significado) e polissemia (quando a mesma palavra tem significados diferentes). A Biblioteca Nacional de Medicina dos Estados Unidos da América (NLM) desenvolveu o Sistema de Linguagem Médica Única (UMLS). Similarmente, um conceito pode ser rastreado até o vocabulário de origem e os termos são mapeados para um ou mais conceitos a níveis mais elevados chamados de tipos semânticos&amp;lt;ref name=&quot;Aronson2001&quot;&gt;Aronson a R: Effective mapping of biomedical text to the UMLS Metathesaurus: the MetaMap program. Proc AMIA Symp 2001:17–21.&amp;lt;/ref&gt;, que podem ser utilizados para resolver a polissemia.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;=== Destilação do Conhecimento ===&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;=== Destilação do Conhecimento ===&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;/table&gt;</summary>
		<author><name>Franzbischoff</name></author>
	</entry>
	<entry>
		<id>https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3570&amp;oldid=prev</id>
		<title>Franzbischoff em 16h55min de 22 de março de 2016</title>
		<link rel="alternate" type="text/html" href="https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3570&amp;oldid=prev"/>
		<updated>2016-03-22T16:55:08Z</updated>

		<summary type="html">&lt;p&gt;&lt;/p&gt;
&lt;table style=&quot;background-color: #fff; color: #202122;&quot; data-mw=&quot;interface&quot;&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;tr class=&quot;diff-title&quot; lang=&quot;pt&quot;&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;← Revisão anterior&lt;/td&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;Revisão das 17h55min de 22 de março de 2016&lt;/td&gt;
				&lt;/tr&gt;&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l7&quot;&gt;Linha 7:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Linha 7:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Embora possa ser tentador comparar o termo &amp;quot;text mining&amp;quot; com [[Data Mining|data mining]], data mining é apenas um dos passos dentro do processo de descoberta de conhecimento. O text mining assemelha-se mais ao processo de descoberta de conhecimento como um todo, englobando os passos de seleção, pré-processamento, transformação, data mining, interpretação e avaliação&amp;lt;ref name=&amp;quot;Blake2011&amp;quot;&amp;gt;Blake C: Text mining. Annu Rev Inf Sci Technol 2011, 45:121–155.&amp;lt;/ref&amp;gt;. É portanto uma tarefa muito mais complexa que o [[Data Mining|data mining]], pois envolve lidar com dados que são inerentemente não estruturados e difusos&amp;lt;ref name=&amp;quot;Tan1999&amp;quot;&amp;gt;Tan A-H: Text Mining: The state of the art and the challenges. Proc PAKDD 1999 Work Knowl Disocovery from Adv Databases 1999, 8:65–70.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Embora possa ser tentador comparar o termo &amp;quot;text mining&amp;quot; com [[Data Mining|data mining]], data mining é apenas um dos passos dentro do processo de descoberta de conhecimento. O text mining assemelha-se mais ao processo de descoberta de conhecimento como um todo, englobando os passos de seleção, pré-processamento, transformação, data mining, interpretação e avaliação&amp;lt;ref name=&amp;quot;Blake2011&amp;quot;&amp;gt;Blake C: Text mining. Annu Rev Inf Sci Technol 2011, 45:121–155.&amp;lt;/ref&amp;gt;. É portanto uma tarefa muito mais complexa que o [[Data Mining|data mining]], pois envolve lidar com dados que são inerentemente não estruturados e difusos&amp;lt;ref name=&amp;quot;Tan1999&amp;quot;&amp;gt;Tan A-H: Text Mining: The state of the art and the challenges. Proc PAKDD 1999 Work Knowl Disocovery from Adv Databases 1999, 8:65–70.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Blake&amp;lt;ref name=&amp;quot;Blake2011&amp;quot;/&amp;gt; define o text mining como a obtenção de padrões novos, interessantes e compreensíveis de uma coleção de texto, enquanto que Feldman&amp;lt;ref name=&amp;quot;Feldman1999&amp;quot;&amp;gt;Feldman R: Mining unstructured data. In Tutorial notes of the fifth ACM SIGKDD international conference on Knowledge discovery and data mining - KDD ’99. New York, New York, USA: ACM Press; 1999:182–236.&amp;lt;/ref&amp;gt; define de uma forma mais semelhante a Recuperação de Informação, onde o objetivo é diferenciar documentos relevantes de documentos irrelevantes. Alguns autores sugerem que o text mining consiste em dois passos: estruturar a informação textual e permitir a descoberta de conhecimento&amp;lt;ref name=&amp;quot;Liddy2000&amp;quot;&amp;gt;Liddy E D: Bulletin Aug/Sept 2000: Interview with Gayle Curtis, Modem Media [http://www.asis.org/Bulletin/Oct-00/liddy.html]&amp;lt;/ref&amp;gt;&amp;lt;ref name=&amp;quot;Nahm2000&amp;quot;&amp;gt;Nahm UY, Mooney RJ: Using Information Extraction to Aid the Discovery of Prediction Rules from Text. In Proceedings of the KDD-2000 Workshop on Text Mining; 2000:51–58.&amp;lt;/ref&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Blake&amp;lt;ref name=&amp;quot;Blake2011&amp;quot;/&amp;gt; define o text mining como a obtenção de padrões novos, interessantes e compreensíveis de uma coleção de texto, enquanto que Feldman&amp;lt;ref name=&amp;quot;Feldman1999&amp;quot;&amp;gt;Feldman R: Mining unstructured data. In Tutorial notes of the fifth ACM SIGKDD international conference on Knowledge discovery and data mining - KDD ’99. New York, New York, USA: ACM Press; 1999:182–236.&amp;lt;/ref&amp;gt; define de uma forma mais semelhante a Recuperação de Informação, onde o objetivo é diferenciar documentos relevantes de documentos irrelevantes. Alguns autores sugerem que o text mining consiste em dois passos: estruturar a informação textual e permitir a descoberta de conhecimento&amp;lt;ref name=&amp;quot;Liddy2000&amp;quot;&amp;gt;Liddy E D: Bulletin Aug/Sept 2000: Interview with Gayle Curtis, Modem Media [http://www.asis.org/Bulletin/Oct-00/liddy.html]&amp;lt;/ref&amp;gt;&amp;lt;ref name=&amp;quot;Nahm2000&amp;quot;&amp;gt;Nahm UY, Mooney RJ: Using Information Extraction to Aid the Discovery of Prediction Rules from Text. In Proceedings of the KDD-2000 Workshop on Text Mining; 2000:51–58.&amp;lt;/ref&amp;gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Acredita-se que cerca de 80% da informação relevante na área dos negócios origina-se de forma não estruturada, principalmente texto&amp;lt;ref name=&amp;quot;Grimes2008&amp;quot;&amp;gt;Grimes S: Unstructured data and the 80 percent rule. Carabridge Bridg 2008:1–2.&amp;lt;/ref&amp;gt;. Na área da saúde, estima-se que 50% da informação que descreve um paciente durante uma terapia é guardada de forma não estruturada como texto livre&amp;lt;ref name=&amp;quot;Sitting2008&amp;quot;&amp;gt;Sittig DF, Wright A, Osheroff J a, Middleton B, Teich JM, Ash JS, Campbell E, Bates DW: Grand challenges in clinical decision support. J Biomed Inform 2008, 41:387–92.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Acredita-se que cerca de 80% da informação relevante na área dos negócios origina-se de forma não estruturada, principalmente texto&amp;lt;ref name=&amp;quot;Grimes2008&amp;quot;&amp;gt;Grimes S: Unstructured data and the 80 percent rule. Carabridge Bridg 2008:1–2.&amp;lt;/ref&amp;gt;. Na área da saúde, estima-se que 50% da informação que descreve um paciente durante uma terapia é guardada de forma não estruturada como texto livre&amp;lt;ref name=&amp;quot;Sitting2008&amp;quot;&amp;gt;Sittig DF, Wright A, Osheroff J a, Middleton B, Teich JM, Ash JS, Campbell E, Bates DW: Grand challenges in clinical decision support. J Biomed Inform 2008, 41:387–92.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;O surgimento de análise de texto na sua forma atual deriva de uma reorientação da investigação nos finais de 1990 do desenvolvimento de algoritmos para aplicações, como descrito pelo Prof Marti A. Hearst o seu artigo ''&amp;quot;Untangling &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;o &lt;/del&gt;Text Data Mining&amp;quot;''&amp;lt;ref name=&amp;quot;Hearst1999&amp;quot;&amp;gt;Hearst M a.: Untangling text data mining. In Proceedings of the 37th annual meeting of the Association for Computational Linguistics on Computational Linguistics -. Morristown, NJ, USA: Association for Computational Linguistics; 1999:3–10.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;O surgimento de análise de texto na sua forma atual deriva de uma reorientação da investigação nos finais de 1990 do desenvolvimento de algoritmos para aplicações, como descrito pelo Prof Marti A. Hearst o seu artigo ''&amp;quot;Untangling Text Data Mining&amp;quot;''&amp;lt;ref name=&amp;quot;Hearst1999&amp;quot;&amp;gt;Hearst M a.: Untangling text data mining. In Proceedings of the 37th annual meeting of the Association for Computational Linguistics on Computational Linguistics -. Morristown, NJ, USA: Association for Computational Linguistics; 1999:3–10.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Processo de Text Mining ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Processo de Text Mining ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;/table&gt;</summary>
		<author><name>Franzbischoff</name></author>
	</entry>
	<entry>
		<id>https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3569&amp;oldid=prev</id>
		<title>Franzbischoff em 16h51min de 22 de março de 2016</title>
		<link rel="alternate" type="text/html" href="https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3569&amp;oldid=prev"/>
		<updated>2016-03-22T16:51:12Z</updated>

		<summary type="html">&lt;p&gt;&lt;/p&gt;
&lt;table style=&quot;background-color: #fff; color: #202122;&quot; data-mw=&quot;interface&quot;&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;tr class=&quot;diff-title&quot; lang=&quot;pt&quot;&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;← Revisão anterior&lt;/td&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;Revisão das 17h51min de 22 de março de 2016&lt;/td&gt;
				&lt;/tr&gt;&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l5&quot;&gt;Linha 5:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Linha 5:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Introdução ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Introdução ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Embora possa ser tentador comparar o termo &amp;quot;text mining&amp;quot; com [[Data Mining|data mining]], data mining é apenas um dos passos dentro do processo de descoberta de conhecimento. O text mining assemelha-se mais ao processo de descoberta de conhecimento como um todo, englobando os passos de seleção, pré-&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;processaento&lt;/del&gt;, transformação, data mining, interpretação e avaliação&amp;lt;ref name=&amp;quot;Blake2011&amp;quot;&amp;gt;Blake C: Text mining. Annu Rev Inf Sci Technol 2011, 45:121–155.&amp;lt;/ref&amp;gt;. É portanto uma tarefa muito mais complexa que o [[Data Mining|data mining]], pois envolve lidar com dados que são inerentemente não estruturados e &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;difuso&lt;/del&gt;&amp;lt;ref name=&amp;quot;Tan1999&amp;quot;&amp;gt;Tan A-H: Text Mining: The state of the art and the challenges. Proc PAKDD 1999 Work Knowl Disocovery from Adv Databases 1999, 8:65–70.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Embora possa ser tentador comparar o termo &amp;quot;text mining&amp;quot; com [[Data Mining|data mining]], data mining é apenas um dos passos dentro do processo de descoberta de conhecimento. O text mining assemelha-se mais ao processo de descoberta de conhecimento como um todo, englobando os passos de seleção, pré-&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;processamento&lt;/ins&gt;, transformação, data mining, interpretação e avaliação&amp;lt;ref name=&amp;quot;Blake2011&amp;quot;&amp;gt;Blake C: Text mining. Annu Rev Inf Sci Technol 2011, 45:121–155.&amp;lt;/ref&amp;gt;. É portanto uma tarefa muito mais complexa que o [[Data Mining|data mining]], pois envolve lidar com dados que são inerentemente não estruturados e &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;difusos&lt;/ins&gt;&amp;lt;ref name=&amp;quot;Tan1999&amp;quot;&amp;gt;Tan A-H: Text Mining: The state of the art and the challenges. Proc PAKDD 1999 Work Knowl Disocovery from Adv Databases 1999, 8:65–70.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Blake&amp;lt;ref name=&amp;quot;Blake2011&amp;quot;/&amp;gt; define o text mining como a obtenção de padrões novos, interessantes e compreensíveis de uma coleção de texto, enquanto que Feldman&amp;lt;ref name=&amp;quot;Feldman1999&amp;quot;&amp;gt;Feldman R: Mining unstructured data. In Tutorial notes of the fifth ACM SIGKDD international conference on Knowledge discovery and data mining - KDD ’99. New York, New York, USA: ACM Press; 1999:182–236.&amp;lt;/ref&amp;gt; define de uma forma mais semelhante a Recuperação de Informação, onde o objetivo é &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;deferenciar &lt;/del&gt;documentos relevantes de documentos irrelevantes. Alguns autores sugerem que o text mining consiste em dois passos: estruturar a informação textual e permitir a descoberta de conhecimento&amp;lt;ref name=&amp;quot;Liddy2000&amp;quot;&amp;gt;Liddy E D: Bulletin Aug/Sept 2000: Interview with Gayle Curtis, Modem Media [http://www.asis.org/Bulletin/Oct-00/liddy.html]&amp;lt;/ref&amp;gt;&amp;lt;ref name=&amp;quot;Nahm2000&amp;quot;&amp;gt;Nahm UY, Mooney RJ: Using Information Extraction to Aid the Discovery of Prediction Rules from Text. In Proceedings of the KDD-2000 Workshop on Text Mining; 2000:51–58.&amp;lt;/ref&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Blake&amp;lt;ref name=&amp;quot;Blake2011&amp;quot;/&amp;gt; define o text mining como a obtenção de padrões novos, interessantes e compreensíveis de uma coleção de texto, enquanto que Feldman&amp;lt;ref name=&amp;quot;Feldman1999&amp;quot;&amp;gt;Feldman R: Mining unstructured data. In Tutorial notes of the fifth ACM SIGKDD international conference on Knowledge discovery and data mining - KDD ’99. New York, New York, USA: ACM Press; 1999:182–236.&amp;lt;/ref&amp;gt; define de uma forma mais semelhante a Recuperação de Informação, onde o objetivo é &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;diferenciar &lt;/ins&gt;documentos relevantes de documentos irrelevantes. Alguns autores sugerem que o text mining consiste em dois passos: estruturar a informação textual e permitir a descoberta de conhecimento&amp;lt;ref name=&amp;quot;Liddy2000&amp;quot;&amp;gt;Liddy E D: Bulletin Aug/Sept 2000: Interview with Gayle Curtis, Modem Media [http://www.asis.org/Bulletin/Oct-00/liddy.html]&amp;lt;/ref&amp;gt;&amp;lt;ref name=&amp;quot;Nahm2000&amp;quot;&amp;gt;Nahm UY, Mooney RJ: Using Information Extraction to Aid the Discovery of Prediction Rules from Text. In Proceedings of the KDD-2000 Workshop on Text Mining; 2000:51–58.&amp;lt;/ref&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Acredita-se que cerca de 80% da informação relevante na área dos negócios origina-se de forma não estruturada, principalmente texto&amp;lt;ref name=&amp;quot;Grimes2008&amp;quot;&amp;gt;Grimes S: Unstructured data and the 80 percent rule. Carabridge Bridg 2008:1–2.&amp;lt;/ref&amp;gt;. Na área da saúde, estima-se que 50% da informação que descreve um paciente durante uma terapia é guardada de forma não estruturada como texto livre&amp;lt;ref name=&amp;quot;Sitting2008&amp;quot;&amp;gt;Sittig DF, Wright A, Osheroff J a, Middleton B, Teich JM, Ash JS, Campbell E, Bates DW: Grand challenges in clinical decision support. J Biomed Inform 2008, 41:387–92.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Acredita-se que cerca de 80% da informação relevante na área dos negócios origina-se de forma não estruturada, principalmente texto&amp;lt;ref name=&amp;quot;Grimes2008&amp;quot;&amp;gt;Grimes S: Unstructured data and the 80 percent rule. Carabridge Bridg 2008:1–2.&amp;lt;/ref&amp;gt;. Na área da saúde, estima-se que 50% da informação que descreve um paciente durante uma terapia é guardada de forma não estruturada como texto livre&amp;lt;ref name=&amp;quot;Sitting2008&amp;quot;&amp;gt;Sittig DF, Wright A, Osheroff J a, Middleton B, Teich JM, Ash JS, Campbell E, Bates DW: Grand challenges in clinical decision support. J Biomed Inform 2008, 41:387–92.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;/table&gt;</summary>
		<author><name>Franzbischoff</name></author>
	</entry>
	<entry>
		<id>https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3568&amp;oldid=prev</id>
		<title>Franzbischoff em 16h48min de 22 de março de 2016</title>
		<link rel="alternate" type="text/html" href="https://aprendis.med.up.pt/index.php?title=Text_Mining&amp;diff=3568&amp;oldid=prev"/>
		<updated>2016-03-22T16:48:27Z</updated>

		<summary type="html">&lt;p&gt;&lt;/p&gt;
&lt;table style=&quot;background-color: #fff; color: #202122;&quot; data-mw=&quot;interface&quot;&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;tr class=&quot;diff-title&quot; lang=&quot;pt&quot;&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;← Revisão anterior&lt;/td&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;Revisão das 17h48min de 22 de março de 2016&lt;/td&gt;
				&lt;/tr&gt;&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l5&quot;&gt;Linha 5:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Linha 5:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Introdução ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Introdução ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Enquanto &lt;/del&gt;o [[Data Mining|data mining]] é &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;aplicado tradicionalmente em coleções &lt;/del&gt;de &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;dados &amp;quot;estruturados&amp;quot;, o &lt;/del&gt;text mining &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;ou text data ining é a aplicação &lt;/del&gt;de &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;técnicas &lt;/del&gt;de data mining &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;em coleções de texto &lt;/del&gt;&amp;quot;&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;não estruturados&amp;quot; ou &amp;quot;semi-estruturados&lt;/del&gt;&amp;quot;. &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;O processo de text &lt;/del&gt;mining &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;tipicamente &lt;/del&gt;envolve &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;o uso de técnicas de processamento de linguagem natural (NLP) para se extrair &lt;/del&gt;dados estruturados &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;de uma narrativa estruturada&lt;/del&gt;&amp;lt;ref name=&amp;quot;&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Zweigenbaum2007&lt;/del&gt;&amp;quot;&amp;gt;&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Zweigenbaum P, Demner&lt;/del&gt;-&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Fushman D, Yu &lt;/del&gt;H&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;, Cohen KB&lt;/del&gt;: &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Frontiers &lt;/del&gt;of &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;biomedical text mining: current progress&lt;/del&gt;. &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Brief Bioinform 2007&lt;/del&gt;, 8:&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;358–75&lt;/del&gt;.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Embora possa ser tentador comparar &lt;/ins&gt;o &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;termo &amp;quot;text mining&amp;quot; com &lt;/ins&gt;[[Data Mining|data mining]]&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;, data mining &lt;/ins&gt;é &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;apenas um dos passos dentro do processo &lt;/ins&gt;de &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;descoberta de conhecimento. O &lt;/ins&gt;text mining &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;assemelha-se mais ao processo de descoberta &lt;/ins&gt;de &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;conhecimento como um todo, englobando os passos &lt;/ins&gt;de &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;seleção, pré-processaento, transformação, &lt;/ins&gt;data mining&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;, interpretação e avaliação&amp;lt;ref name=&lt;/ins&gt;&amp;quot;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Blake2011&lt;/ins&gt;&amp;quot;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&amp;gt;Blake C: Text mining. Annu Rev Inf Sci Technol 2011, 45:121–155.&amp;lt;/ref&amp;gt;&lt;/ins&gt;. &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;É portanto uma tarefa muito mais complexa que o [[Data Mining|data &lt;/ins&gt;mining&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;]], pois &lt;/ins&gt;envolve &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;lidar com &lt;/ins&gt;dados &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;que são inerentemente não &lt;/ins&gt;estruturados &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;e difuso&lt;/ins&gt;&amp;lt;ref name=&amp;quot;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Tan1999&lt;/ins&gt;&amp;quot;&amp;gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Tan A&lt;/ins&gt;-H: &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Text Mining: The state &lt;/ins&gt;of &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;the art and the challenges&lt;/ins&gt;. &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Proc PAKDD 1999 Work Knowl Disocovery from Adv Databases 1999&lt;/ins&gt;, 8:&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;65–70&lt;/ins&gt;.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;O &lt;/del&gt;text mining &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;inicia por extrair factos &lt;/del&gt;e &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;eventos &lt;/del&gt;de &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;fontes &lt;/del&gt;de texto, &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;permitindo a formação de hipóteses &lt;/del&gt;que &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;são depois exploradas por métodos tradicionais de análise de dados&lt;/del&gt;&amp;lt;ref name=&amp;quot;&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Maletic2010&lt;/del&gt;&amp;quot;&amp;gt;&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Maletic JI, Marcus A&lt;/del&gt;: &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Data &lt;/del&gt;Mining and &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Knowledge Discovery Handbook&lt;/del&gt;. &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;2nd edition. Boston&lt;/del&gt;, &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;MA&lt;/del&gt;: &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Springer US&lt;/del&gt;; &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;2010&lt;/del&gt;.&amp;lt;/ref&amp;gt;. &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Ele pode ser dividido &lt;/del&gt;em &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;duas etapas principais&lt;/del&gt;: &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Refinação &lt;/del&gt;de &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Texto que transforma o texto livre em um formato intermediário&lt;/del&gt;, &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;tal como um gráfico conceptual (semi&lt;/del&gt;-&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;estruturado) ou dados relacionados (estruturado), e a Destilação do Conhecimento&lt;/del&gt;, que &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;deduz padrões ou conhecimento a partir desta &lt;/del&gt;forma &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;intermediária&lt;/del&gt;. &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;O formato intermediário pode ser baseado em documentos&lt;/del&gt;, &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;em &lt;/del&gt;que &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;cada entidade representa um documento, ou baseado em conceitos, em &lt;/del&gt;que &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;cada entidade representa &lt;/del&gt;um &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;objeto ou conceito &lt;/del&gt;de &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;interesse&lt;/del&gt;&amp;lt;ref name=&amp;quot;&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Tan1999&lt;/del&gt;&amp;quot;&amp;gt;&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Tan &lt;/del&gt;A&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;-H&lt;/del&gt;: Text Mining: &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;The state &lt;/del&gt;of the &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;art and &lt;/del&gt;the &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;challenges&lt;/del&gt;. &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Proc PAKDD 1999 Work Knowl Disocovery from Adv Databases &lt;/del&gt;1999&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;, 8&lt;/del&gt;:&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;65–70&lt;/del&gt;.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Blake&amp;lt;ref name=&amp;quot;Blake2011&amp;quot;/&amp;gt; define o &lt;/ins&gt;text mining &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;como a obtenção de padrões novos, interessantes &lt;/ins&gt;e &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;compreensíveis &lt;/ins&gt;de &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;uma coleção &lt;/ins&gt;de texto, &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;enquanto &lt;/ins&gt;que &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Feldman&lt;/ins&gt;&amp;lt;ref name=&amp;quot;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Feldman1999&lt;/ins&gt;&amp;quot;&amp;gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Feldman R&lt;/ins&gt;: Mining &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;unstructured data. In Tutorial notes of the fifth ACM SIGKDD international conference on Knowledge discovery &lt;/ins&gt;and &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;data mining - KDD ’99&lt;/ins&gt;. &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;New York, New York&lt;/ins&gt;, &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;USA&lt;/ins&gt;: &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;ACM Press&lt;/ins&gt;; &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;1999:182–236&lt;/ins&gt;.&amp;lt;/ref&amp;gt; &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;define de uma forma mais semelhante a Recuperação de Informação, onde o objetivo é deferenciar documentos relevantes de documentos irrelevantes&lt;/ins&gt;. &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Alguns autores sugerem que o text mining consiste &lt;/ins&gt;em &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;dois passos&lt;/ins&gt;: &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;estruturar a informação textual e permitir a descoberta &lt;/ins&gt;de &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;conhecimento&amp;lt;ref name=&amp;quot;Liddy2000&amp;quot;&amp;gt;Liddy E D: Bulletin Aug/Sept 2000: Interview with Gayle Curtis&lt;/ins&gt;, &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Modem Media [http://www.asis.org/Bulletin/Oct&lt;/ins&gt;-&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;00/liddy.html]&amp;lt;/ref&amp;gt;&amp;lt;ref name=&amp;quot;Nahm2000&amp;quot;&amp;gt;Nahm UY&lt;/ins&gt;, &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Mooney RJ: Using Information Extraction to Aid the Discovery of Prediction Rules from Text. In Proceedings of the KDD-2000 Workshop on Text Mining; 2000:51–58.&amp;lt;/ref&amp;gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt; &lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Acredita-se &lt;/ins&gt;que &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;cerca de 80% da informação relevante na área dos negócios origina-se de &lt;/ins&gt;forma &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;não estruturada, principalmente texto&amp;lt;ref name=&amp;quot;Grimes2008&amp;quot;&amp;gt;Grimes S: Unstructured data and the 80 percent rule. Carabridge Bridg 2008:1–2&lt;/ins&gt;.&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&amp;lt;/ref&amp;gt;. Na área da saúde&lt;/ins&gt;, &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;estima-se &lt;/ins&gt;que &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;50% da informação &lt;/ins&gt;que &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;descreve &lt;/ins&gt;um &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;paciente durante uma terapia é guardada &lt;/ins&gt;de &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;forma não estruturada como texto livre&lt;/ins&gt;&amp;lt;ref name=&amp;quot;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Sitting2008&lt;/ins&gt;&amp;quot;&amp;gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Sittig DF, Wright &lt;/ins&gt;A&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;, Osheroff J a, Middleton B, Teich JM, Ash JS, Campbell E, Bates DW: Grand challenges in clinical decision support. J Biomed Inform 2008, 41&lt;/ins&gt;:&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;387–92.&amp;lt;/ref&amp;gt;.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt; &lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;O surgimento de análise de texto na sua forma atual deriva de uma reorientação da investigação nos finais de 1990 do desenvolvimento de algoritmos para aplicações, como descrito pelo Prof Marti A. Hearst o seu artigo ''&amp;quot;Untangling o &lt;/ins&gt;Text &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Data &lt;/ins&gt;Mining&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&amp;quot;''&amp;lt;ref name=&amp;quot;Hearst1999&amp;quot;&amp;gt;Hearst M a.&lt;/ins&gt;: &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Untangling text data mining. In Proceedings &lt;/ins&gt;of the &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;37th annual meeting of &lt;/ins&gt;the &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Association for Computational Linguistics on Computational Linguistics -&lt;/ins&gt;. &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Morristown, NJ, USA: Association for Computational Linguistics; &lt;/ins&gt;1999:&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;3–10&lt;/ins&gt;.&amp;lt;/ref&amp;gt;.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Processo de Text Mining ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Processo de Text Mining ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Enquanto o [[Data Mining|data mining]] é aplicado tradicionalmente em coleções de dados &quot;estruturados&quot;, o text mining ou text data mining é a aplicação de técnicas de [[Data Mining|data mining]] em coleções de texto &quot;não estruturados&quot; ou &quot;semi-estruturados&quot;. O processo de text mining tipicamente envolve o uso de técnicas de processamento de linguagem natural (NLP) para se extrair dados estruturados de uma narrativa estruturada&amp;lt;ref name=&quot;Zweigenbaum2007&quot;&gt;Zweigenbaum P, Demner-Fushman D, Yu H, Cohen KB: Frontiers of biomedical text mining: current progress. Brief Bioinform 2007, 8:358–75.&amp;lt;/ref&gt;.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;O text mining inicia por extrair factos e eventos de fontes de texto, permitindo a formação de hipóteses que são depois exploradas por métodos tradicionais de análise de dados&amp;lt;ref name=&quot;Maletic2010&quot;&gt;Maletic JI, Marcus A: Data Mining and Knowledge Discovery Handbook. 2nd edition. Boston, MA: Springer US; 2010.&amp;lt;/ref&gt;. Ele pode ser dividido em duas etapas principais: Refinação de Texto que transforma o texto livre em um formato intermediário, tal como um gráfico conceptual (semi-estruturado) ou dados relacionados (estruturado), e a Destilação do Conhecimento, que deduz padrões ou conhecimento a partir desta forma intermediária. O formato intermediário pode ser baseado em documentos, em que cada entidade representa um documento, ou baseado em conceitos, em que cada entidade representa um objeto ou conceito de interesse&amp;lt;ref name=&quot;Tan1999&quot;/&gt;.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br/&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;=== Refinação do Texto ===&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;=== Refinação do Texto ===&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;/table&gt;</summary>
		<author><name>Franzbischoff</name></author>
	</entry>
</feed>