You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

397 lines
25 KiB
HTML

This file contains ambiguous Unicode characters!

This file contains ambiguous Unicode characters that may be confused with others in your current locale. If your use case is intentional and legitimate, you can safely ignore this warning. Use the Escape button to highlight these characters.

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html>
<head>
<title>pattern</title>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<link type="text/css" rel="stylesheet" href="../clips.css" />
<style>
/* Small fixes because we omit the online layout.css. */
h3 { line-height: 1.3em; }
#page { margin-left: auto; margin-right: auto; }
#header, #header-inner { height: 175px; }
#header { border-bottom: 1px solid #C6D4DD; }
table { border-collapse: collapse; }
#checksum { display: none; }
</style>
<link href="../js/shCore.css" rel="stylesheet" type="text/css" />
<link href="../js/shThemeDefault.css" rel="stylesheet" type="text/css" />
<script language="javascript" src="../js/shCore.js"></script>
<script language="javascript" src="../js/shBrushXml.js"></script>
<script language="javascript" src="../js/shBrushJScript.js"></script>
<script language="javascript" src="../js/shBrushPython.js"></script>
</head>
<body class="node-type-page one-sidebar sidebar-right section-pages">
<div id="page">
<div id="page-inner">
<div id="header"><div id="header-inner"></div></div>
<div id="content">
<div id="content-inner">
<div class="node node-type-page"
<div class="node-inner">
<div class="breadcrumb">View online at: <a href="http://www.clips.ua.ac.be/pages/pattern" class="noexternal" target="_blank">http://www.clips.ua.ac.be/pages/pattern</a></div>
<h1>pattern</h1>
<!-- Parsed from the online documentation. -->
<div id="node-1350" class="node node-type-page"><div class="node-inner">
<div class="content">
<p><span class="big">Pattern is a web mining module for the Python programming language.</span></p>
<p><span class="big">It has tools for data mining (Google, Twitter and Wikipedia API, a web crawler, a HTML DOM parser), natural language processing (part-of-speech taggers, n-gram search, sentiment analysis, WordNet), machine learning (vector space model, clustering, SVM), network analysis and &lt;canvas&gt; visualization.</span></p>
<p>The module is free, well-document and bundled with 50+ examples and 350+ unit tests.</p>
<p><img src="../g/pattern_schema.gif" alt="" width="620" height="180" /></p>
<hr />
<h2>Download</h2>
<table>
<tbody>
<tr>
<td><a onclick="javascript:_gaq.push(['_trackPageview', '/downloads/pattern']);" href="http://www.clips.ua.ac.be/media/pattern-2.6.zip" target="_self"><img src="../g/download.gif" alt="download" align="left" /></a></td>
<td><strong>Pattern 2.6</strong>&nbsp;| <a onclick="javascript:_gaq.push(['_trackPageview', '/downloads/pattern']);" href="http://www.clips.ua.ac.be/media/pattern-2.6.zip" target="_self">download</a> (.zip, 23MB)<br />
<ul>
<li>Requires: Python 2.5+ on Windows | Mac | Linux</li>
<li>Licensed under <a href="http://www.linfo.org/bsdlicense.html" target="_blank">BSD</a></li>
<li>Latest releases: <a class="noexternal" href="http://www.clips.ua.ac.be/media/pattern-2.6.zip">2.6</a> |&nbsp;<a class="noexternal" href="http://www.clips.ua.ac.be/media/pattern-2.5.zip">2.5</a> |&nbsp;<a class="noexternal" href="http://www.clips.ua.ac.be/media/pattern-2.4.zip">2.4</a> | <a class="noexternal" href="http://www.clips.ua.ac.be/media/pattern-2.3.zip">2.3</a>&nbsp;|&nbsp;<a class="noexternal" href="http://www.clips.ua.ac.be/media/pattern-2.2.zip">2.2</a> |&nbsp;<a class="noexternal" href="http://www.clips.ua.ac.be/media/pattern-2.1.zip">2.1</a> |&nbsp;<a class="noexternal" href="http://www.clips.ua.ac.be/media/pattern-2.0.zip">2.0</a></li>
<li>Authors:<br />&nbsp;Tom De Smedt (<em>tom at organisms.be</em>)<br />&nbsp;Walter Daelemans&nbsp;</li>
</ul>
<p><span class="small"><span style="text-decoration: underline;">Reference</span>: De Smedt, T. &amp; Daelemans, W. (2012)</span>.<br /><span class="small">Pattern for Python. <em>Journal of Machine Learning Research</em>, 13: 20312035.</span></p>
<p id="checksum" class="grey"><span class="small"><span style="text-decoration: underline;">SHA256</span> checksum&nbsp;of the .zip:<br />28213f05d94a86d2de1d8a03525d456a9e68dc3b563dc2481ad08fe3db180d02</span></p>
</td>
<td>
</td>
</tr>
</tbody>
</table>
<p>&nbsp;</p>
<hr />
<table border="0">
<tbody>
<tr>
<td style="width: 200px;">
<h2>Modules</h2>
<ul>
<li><a href="pattern-web.html">pattern.web</a></li>
<li><a href="pattern-db.html">pattern.db</a></li>
<li><a href="pattern-en.html">pattern.en</a>&nbsp;|&nbsp;<a href="pattern-es.html">es</a>&nbsp;| <a href="pattern-de.html">de</a> | <a href="pattern-fr.html">fr</a> | <a href="pattern-it.html">it</a> |&nbsp;<a href="pattern-nl.html">nl</a></li>
<li><a href="pattern-search.html">pattern.search</a></li>
<li><a href="pattern-vector.html">pattern.vector</a></li>
<li><a href="pattern-graph.html">pattern.graph</a>&nbsp;</li>
</ul>
<p><span class="smallcaps">Helper modules</span></p>
<ul style="margin-top: 0;">
<li><a href="pattern-metrics.html">pattern.metrics</a></li>
<li><a href="pattern-canvas.html">canvas.js</a></li>
</ul>
<p><span class="smallcaps">Command-line</span></p>
<ul style="margin-top: 0;">
<li><a href="pattern-shell.html">Command-line interface</a></li>
</ul>
</td>
<td>
<h2><a name="contribute"></a>Contribute</h2>
<ul>
<li><a href="pattern-dev.html">Developer documentation</a></li>
<li><a href="https://github.com/clips/pattern" target="_blank">GitHub repository</a></li>
<li><a href="http://groups.google.com/group/pattern-for-python" target="_blank">Google group</a></li>
</ul>
<form action="https://www.paypal.com/cgi-bin/webscr" method="post"><input type="hidden" name="cmd" value="_s-xclick" /> <input type="hidden" name="hosted_button_id" value="HW2GU5PNWYQV8" /> <input type="image" name="submit" src="../g/paypal-donate.jpg" alt="PayPal - The safer, easier way to pay online!" /> <img src="https://www.paypalobjects.com/en_US/i/scr/pixel.gif" alt="" width="1" height="1" border="0" /></form>
</td>
</tr>
</tbody>
</table>
<p>&nbsp;</p>
<hr />
<h2>Installation</h2>
<p>Pattern is written for Python 2.5+ (also supports Python 3.6+). The module has no external dependencies, except <span class="inline_code">LSA</span> in the pattern.vector module, which requires <a href="http://numpy.scipy.org/" target="_blank">NumPy</a> (installed by default on Mac OS X).&nbsp;</p>
<p>To install Pattern so that the module is available in all Python scripts, from the command line do:</p>
<div class="install">
<pre class="gutter:false; light:true;">&gt; cd pattern-3.6
&gt; python setup.py install&nbsp;</pre></div>
<p>If you have pip, you can automatically download and install from the PyPi repository:</p>
<div class="install">
<pre class="gutter:false; light:true;">&gt; pip install pattern</pre></div>
<p>If none of the above works, you can make Python aware of the module in three ways:</p>
<ul>
<li>Put the <span class="inline_code">pattern</span>&nbsp;subfolder in the .zip archive in the same folder as your script.</li>
<li>Put the <span class="inline_code">pattern</span>&nbsp;subfolder in the standard location for modules so it is available to all scripts:<br /><span class="inline_code">c:\python27\Lib\site-packages\</span>&nbsp;(Windows),<br /><span class="inline_code"> /Library/Python/2.7/site-packages/</span>&nbsp;(Mac),<br /><span class="inline_code">/usr/lib/python2.7/site-packages/</span>&nbsp;(Unix).<span style="font-family: Courier, monospace; font-size: small;"><span style="font-size: 12px;"><br /></span></span></li>
<li>Add the location of the module to&nbsp;<span class="inline_code">sys.path</span>&nbsp;in your Python script, before importing it:</li>
</ul>
<div class="example">
<pre class="brush:python; gutter:false; light:true;">&gt;&gt;&gt; import sys; sys.path.append('/users/tom/desktop/pattern')
&gt;&gt;&gt; from pattern.web import Twitter </pre></div>
<p>&nbsp;</p>
<hr />
<h2>Quick overview</h2>
<h3>pattern.web</h3>
<p>The&nbsp;<a href="pattern-web.html">pattern.web</a>&nbsp;module is a web toolkit that contains API's (Google, Gmail, Bing, Twitter, Facebook, Wikipedia, Wiktionary, DBPedia, Flickr, ...), a robust HTML DOM parser and a web crawler.</p>
<div class="example">
<pre class="brush:python; gutter:false; light:true;">&gt;&gt;&gt; from pattern.web import Twitter, plaintext
&gt;&gt;&gt;
&gt;&gt;&gt; twitter = Twitter(language='en')
&gt;&gt;&gt; for tweet in twitter.search('"more important than"', cached=False):
&gt;&gt;&gt; print plaintext(tweet.text)
'The mobile web is more important than mobile apps.'
'Start slowly, direction is more important than speed.'
'Imagination is more important than knowledge. - Albert Einstein'
... </pre></div>
<h3>pattern.en</h3>
<p>The&nbsp;<a href="pattern-en.html">pattern.en</a>&nbsp;module is a natural language processing (NLP) toolkit for English. Because language is ambiguous (e.g., <em>I can</em>&nbsp;<em>a can</em>) it uses statistical approaches + regular expressions. This means that it is fast, quite accurate and occasionally incorrect. It has a part-of-speech tagger that identifies word types (e.g., noun, verb, adjective), word inflection (conjugation, singularization) and a WordNet API.</p>
<div class="example">
<pre class="brush:python; gutter:false; light:true;">&gt;&gt;&gt; from pattern.en import parse
&gt;&gt;&gt;
&gt;&gt;&gt; s = 'The mobile web is more important than mobile apps.'
&gt;&gt;&gt; s = parse(s, relations=True, lemmata=True)
&gt;&gt;&gt; print s
'The/DT/B-NP/O/NP-SBJ-1/the mobile/JJ/I-NP/O/NP-SBJ-1/mobile' ...
</pre></div>
<table class="border">
<tbody>
<tr>
<td class="smallcaps" style="text-align: right;">word</td>
<td class="smallcaps" style="text-align: center;">tag</td>
<td class="smallcaps" style="text-align: center;">chunk</td>
<td class="smallcaps" style="text-align: center;">role</td>
<td class="smallcaps" style="text-align: center;">id</td>
<td class="smallcaps" style="text-align: center;">pnp</td>
<td class="smallcaps">lemma</td>
</tr>
<tr>
<td style="text-align: right;">The</td>
<td class="inline_code" style="text-align: center;">DT</td>
<td class="inline_code" style="text-align: center;">NP&nbsp;</td>
<td class="inline_code" style="text-align: center;">SBJ</td>
<td class="inline_code" style="text-align: center;">1</td>
<td class="inline_code" style="text-align: center;">-</td>
<td><em>the</em></td>
</tr>
<tr>
<td style="text-align: right;">mobile</td>
<td class="inline_code" style="text-align: center;">JJ</td>
<td class="inline_code" style="text-align: center;">NP^</td>
<td class="inline_code" style="text-align: center;">SBJ</td>
<td class="inline_code" style="text-align: center;">1</td>
<td class="inline_code" style="text-align: center;">-</td>
<td><em>mobile</em></td>
</tr>
<tr>
<td style="text-align: right;">web</td>
<td class="inline_code" style="text-align: center;">NN</td>
<td class="inline_code" style="text-align: center;">NP^</td>
<td class="inline_code" style="text-align: center;">SBJ</td>
<td class="inline_code" style="text-align: center;">1</td>
<td class="inline_code" style="text-align: center;">-</td>
<td><em>web</em></td>
</tr>
<tr>
<td style="text-align: right;">is</td>
<td class="inline_code" style="text-align: center;">VBZ</td>
<td class="inline_code" style="text-align: center;">VP&nbsp;</td>
<td class="inline_code" style="text-align: center;">-</td>
<td class="inline_code" style="text-align: center;">1</td>
<td class="inline_code" style="text-align: center;">-</td>
<td><em>be</em></td>
</tr>
<tr>
<td style="text-align: right;">more</td>
<td class="inline_code" style="text-align: center;">RBR</td>
<td class="inline_code" style="text-align: center;">ADJP&nbsp;</td>
<td class="inline_code" style="text-align: center;">-</td>
<td class="inline_code" style="text-align: center;">-</td>
<td class="inline_code" style="text-align: center;">-</td>
<td><em>more</em></td>
</tr>
<tr>
<td style="text-align: right;">important</td>
<td class="inline_code" style="text-align: center;">JJ</td>
<td class="inline_code" style="text-align: center;">ADJP^</td>
<td class="inline_code" style="text-align: center;">-</td>
<td class="inline_code" style="text-align: center;">-</td>
<td class="inline_code" style="text-align: center;">-</td>
<td><em>important</em></td>
</tr>
<tr>
<td style="text-align: right;">than</td>
<td class="inline_code" style="text-align: center;">IN</td>
<td class="inline_code" style="text-align: center;">PP&nbsp;</td>
<td class="inline_code" style="text-align: center;">-</td>
<td class="inline_code" style="text-align: center;">-</td>
<td class="inline_code" style="text-align: center;">PNP</td>
<td><em>than</em></td>
</tr>
<tr>
<td style="text-align: right;">mobile</td>
<td class="inline_code" style="text-align: center;">JJ</td>
<td class="inline_code" style="text-align: center;">NP&nbsp;</td>
<td class="inline_code" style="text-align: center;">-</td>
<td class="inline_code" style="text-align: center;">-</td>
<td class="inline_code" style="text-align: center;">PNP</td>
<td><em>mobile</em></td>
</tr>
<tr>
<td style="text-align: right;">apps</td>
<td class="inline_code" style="text-align: center;">NNS</td>
<td class="inline_code" style="text-align: center;">NP^</td>
<td class="inline_code" style="text-align: center;">-</td>
<td class="inline_code" style="text-align: center;">-</td>
<td class="inline_code" style="text-align: center;">PNP</td>
<td><em>app</em></td>
</tr>
<tr>
<td style="text-align: right;">.</td>
<td class="inline_code" style="text-align: center;">.</td>
<td class="inline_code" style="text-align: center;">-&nbsp;</td>
<td class="inline_code" style="text-align: center;">-</td>
<td class="inline_code" style="text-align: center;">-</td>
<td class="inline_code" style="text-align: center;">-</td>
<td>.</td>
</tr>
</tbody>
</table>
<p>The text has been annotated with word types,&nbsp;for example nouns (<span class="postag">NN</span>), verbs(<span class="postag">VB</span>),&nbsp;adjectives (<span class="postag">JJ</span>) and determiners (<span class="postag">DT</span>), word types (e.g.,&nbsp;sentence subject&nbsp;<span class="postag">SBJ</span>) and prepositional noun phrases (<span class="postag">PNP</span>). To iterate over the parts in the tagged text we can construct a <em>parse tree</em>.</p>
<div class="example">
<pre class="brush:python; gutter:false; light:true;">&gt;&gt;&gt; from pattern.en import parsetree
&gt;&gt;&gt;
&gt;&gt;&gt; s = 'The mobile web is more important than mobile apps.'
&gt;&gt;&gt; s = parsetree(s)
&gt;&gt;&gt; for sentence in s:
&gt;&gt;&gt; for chunk in sentence.chunks:
&gt;&gt;&gt; for word in chunk.words:
&gt;&gt;&gt; print word,
&gt;&gt;&gt; print
Word(u'The/DT') Word(u'mobile/JJ') Word(u'web/NN')
Word(u'is/VBZ')
Word(u'more/RBR') Word(u'important/JJ')
Word(u'than/IN')
Word(u'mobile/JJ') Word(u'apps/NNS')
</pre></div>
<p>Parsers for Spanish, French, Italian, German and Dutch are also available: <br /><a href="pattern-es.html">pattern.es</a>&nbsp;| <a href="pattern-fr.html">pattern.fr</a> | <a href="pattern-it.html">pattern.it</a> |&nbsp;<a href="pattern-de.html">pattern.de</a>&nbsp;|&nbsp;<a href="pattern-nl.html">pattern.nl</a></p>
<h3>pattern.search</h3>
<p>The&nbsp;<a href="pattern-search.html">pattern.search</a>&nbsp;module contains a search algorithm to retrieve sequences of words (called <em>n-grams</em>) from tagged text.</p>
<div class="example">
<pre class="brush:python; gutter:false; light:true;">&gt;&gt;&gt; from pattern.en import parsetree
&gt;&gt;&gt; from pattern.search import search
&gt;&gt;&gt;
&gt;&gt;&gt; s = 'The mobile web is more important than mobile apps.'
&gt;&gt;&gt; s = parsetree(s, relations=True, lemmata=True)
&gt;&gt;&gt;
&gt;&gt;&gt; for match in search('NP be RB?+ important than NP', s):
&gt;&gt;&gt; print match.constituents()[-1], '=&gt;', \
&gt;&gt;&gt; match.constituents()[0]
Chunk('mobile apps/NP') =&gt; Chunk('The mobile web/NP-SBJ-1')
</pre></div>
<p>The search pattern&nbsp;<span class="inline_code">NP</span> <span class="inline_code">be</span> <span class="inline_code">RB?+</span> <span class="inline_code">important</span> <span class="inline_code">than</span> <span class="inline_code">NP</span> means any noun phrase (<span class="postag">NP</span>) followed by the verb <em>to be</em>, followed by zero or more adverbs (<span class="postag">RB</span>, e.g.,&nbsp;<em>much</em>, <em>more</em>), followed by the words <em>important than</em>, followed by any noun phrase. It will also match "<em>The mobile web <span style="text-decoration: underline;">will</span> <span style="text-decoration: underline;">be</span> <span style="text-decoration: underline;">much</span> <span style="text-decoration: underline;">less</span> important than mobile apps</em>" and other grammatical variations.</p>
<h3>pattern.vector</h3>
<p>The&nbsp;<a href="pattern-vector.html">pattern.vector</a>&nbsp;module is a toolkit for machine learning, based on a vector space model&nbsp;of bag-of-words documents with weighted features (e.g., tf-idf) and distance metrics (e.g., cosine similarity, infogain).&nbsp;Models can be used for clustering (<em>k</em>-means, hierarchical), classification (Naive Bayes, Perceptron,&nbsp;<em>k-</em>NN, SVM) and latent semantic analysis (LSA).</p>
<div>
<div class="example">
<pre class="brush: python;gutter: false; fontsize: 100; first-line: 1; ">&gt;&gt;&gt; from pattern.web import Twitter
&gt;&gt;&gt; from pattern.en import tag
&gt;&gt;&gt; from pattern.vector import KNN, count
&gt;&gt;&gt;
&gt;&gt;&gt; twitter, knn = Twitter(), KNN()
&gt;&gt;&gt;
&gt;&gt;&gt; for i in range(1, 10):
&gt;&gt;&gt; for tweet in twitter.search('#win OR #fail', start=i, count=100):
&gt;&gt;&gt; s = tweet.text.lower()
&gt;&gt;&gt; p = '#win' in s and 'WIN' or 'FAIL'
&gt;&gt;&gt; v = tag(s)
&gt;&gt;&gt; v = [word for word, pos in v if pos == 'JJ'] # JJ = adjective
&gt;&gt;&gt; v = count(v)
&gt;&gt;&gt; if v:
&gt;&gt;&gt; knn.train(v, type=p)
&gt;&gt;&gt;
&gt;&gt;&gt; print knn.classify('sweet potato burger')
&gt;&gt;&gt; print knn.classify('stupid autocorrect')
'WIN'
'FAIL' </pre></div>
</div>
<p>This example trains a classifier on adjectives mined from Twitter. First, tweets with hashtag #win or #fail are mined. For example: <em>"$20 tip off a <span style="text-decoration: underline;">sweet</span> <span style="text-decoration: underline;">little</span> <span style="text-decoration: underline;">old</span> lady today #win"</em>. The word part-of-speech tags are parsed, keeping only adjectives. Each tweet is transformed to a vector, a dictionary of adjective → count items, labeled <span class="inline_code">WIN</span> or <span class="inline_code">FAIL</span>. The classifier uses the vectors to learn which other, unknown tweets look more like&nbsp;<span class="inline_code">WIN</span>&nbsp;(e.g., <em>sweet potato burger</em>) or more like <span class="inline_code">FAIL</span> (e.g., <em>stupid autocorrect</em>).</p>
<h3>pattern.graph</h3>
<p>The&nbsp;<a href="pattern-graph.html">pattern.graph</a>&nbsp;module provides a graph data structure that represents relations between nodes (e.g., terms, concepts). Graphs can be exported as HTML <span class="inline_code">&lt;canvas&gt;</span> animations (<span class="link-maintenance"><a href="http://www.clips.ua.ac.be/media/pattern-graph" target="_blank">demo</a></span>). In the example below, more <em>central</em> nodes (= more incoming traffic) are colored in blue.</p>
<p><img class="border" src="../g/pattern_graph5.jpg" alt="" width="610" height="198" /></p>
<div class="example">
<pre class="brush:python; gutter:false; light:true;">&gt;&gt;&gt; from pattern.web import Bing, plaintext
&gt;&gt;&gt; from pattern.en import parsetree
&gt;&gt;&gt; from pattern.search import search
&gt;&gt;&gt; from pattern.graph import Graph
&gt;&gt;&gt;
&gt;&gt;&gt; g = Graph()
&gt;&gt;&gt; for i in range(10):
&gt;&gt;&gt; for result in Bing().search('"more important than"', start=i+1, count=50):
&gt;&gt;&gt; s = r.text.lower()
&gt;&gt;&gt; s = plaintext(s)
&gt;&gt;&gt; s = parsetree(s)
&gt;&gt;&gt; p = '{NP} (VP) more important than {NP}'
&gt;&gt;&gt; for m in search(p, s):
&gt;&gt;&gt; x = m.group(1).string # NP left
&gt;&gt;&gt; y = m.group(2).string # NP right
&gt;&gt;&gt; if x not in g:
&gt;&gt;&gt; g.add_node(x)
&gt;&gt;&gt; if y not in g:
&gt;&gt;&gt; g.add_node(y)
&gt;&gt;&gt; g.add_edge(g[x], g[y], stroke=(0,0,0,0.75)) # R,G,B,A
&gt;&gt;&gt;
&gt;&gt;&gt; g = g.split()[0] # Largest subgraph.
&gt;&gt;&gt;
&gt;&gt;&gt; for n in g.sorted()[:40]: # Sort by Node.weight.
&gt;&gt;&gt; n.fill = (0, 0.5, 1, 0.75 * n.weight)
&gt;&gt;&gt;
&gt;&gt;&gt; g.export('test', directed=True, weighted=0.6) </pre></div>
<p>Some relations (= edges) could use some extra post-processing, e.g., in <em>nothing is more important than life</em>, <em>nothing</em> is <span style="text-decoration: underline;">not</span> more important than <em>life</em>.</p>
<p>&nbsp;</p>
<hr />
<h2>Case studies&nbsp;</h2>
<p>Case studies with hands-on source code examples.</p>
<table border="0">
<tbody>
<tr>
<td>
<p><a href="http://www.clips.ua.ac.be/pages/modeling-creativity-with-a-semantic-network-of-common-sense"><img src="../g/pattern_example_semantic_network.jpg" alt="" width="70" height="70" /><br /></a></p>
</td>
<td>&nbsp;</td>
<td><span class="smallcaps">modeling creativity with a semantic network of common sense </span><span class="small">(2013)</span>&nbsp;<br />This case study offers a computational model of creativity, by representing the mind as a semantic network of common sense, using <a class="link-maintenance" href="pattern-graph.html">pattern.graph</a>&nbsp;&amp; <a class="link-maintenance" href="pattern-web.html">web</a>.<br /><a href="http://www.clips.ua.ac.be/pages/modeling-creativity-with-a-semantic-network-of-common-sense">read more »</a></td>
</tr>
<tr>
<td>
<p><a class="noexternal" href="http://www.clips.ua.ac.be/pages/using-wiktionary-to-build-an-italian-part-of-speech-tagger"><img src="../g/pattern_example_italian.jpg" alt="" width="70" height="70" /><br /></a></p>
</td>
<td>&nbsp;</td>
<td><span class="smallcaps">using wiktionary to build an italian part-of-speech tagger </span><span class="small">(2013)</span> <br />This case study demonstrates how a part-of-speech tagger for Italian (see <a class="link-maintenance" href="pattern-it.html">pattern.it</a>) can be built by mining Wiktionary and Wikipedia. &nbsp;<br /><a href="http://www.clips.ua.ac.be/pages/using-wiktionary-to-build-an-italian-part-of-speech-tagger">read more »</a></td>
</tr>
<tr>
<td>
<p><a class="noexternal" href="http://www.clips.ua.ac.be/pages/using-wikicorpus-nltk-to-build-a-spanish-part-of-speech-tagger"><img src="../g/pattern_example_spanish.jpg" alt="" width="70" height="70" /><br /></a></p>
</td>
<td>&nbsp;</td>
<td><span class="smallcaps">using wikicorpus and nltk to build a spanish part-of-speech tagger </span><span class="small">(2012)</span><br />This case study demonstrates how a part-of-speech tagger for Spanish (see <a class="link-maintenance" href="pattern-es.html">pattern.es</a>) can be built by using NLTK and the freely available Wikicorpus. <br /><a href="http://www.clips.ua.ac.be/pages/using-wikicorpus-nltk-to-build-a-spanish-part-of-speech-tagger">read more »</a></td>
</tr>
<tr>
<td>
<p><a class="noexternal" href="http://www.clips.ua.ac.be/pages/pattern-examples-elections"><img src="../g/pattern_example_elections.jpg" alt="" width="70" height="70" /><br /></a></p>
</td>
<td>&nbsp;</td>
<td><span class="smallcaps">belgian elections</span><span class="smallcaps">, twitter sentiment analysis&nbsp;</span><span class="small">(2010)</span><br />This case study uses sentiment analysis (e.g., positive or negative tone) on 7,500 Dutch and French tweets (see <a class="link-maintenance" href="pattern-web.html">pattern.web</a> |&nbsp;<a class="link-maintenance" href="pattern-nl.html">nl</a>&nbsp;|&nbsp;<a class="link-maintenance" href="pattern-fr.html">fr</a>) in the weeks before the Belgian 2010 elections. <br /><a href="http://www.clips.ua.ac.be/pages/pattern-examples-elections">read more »</a></td>
</tr>
<tr>
<td>
<p><a class="noexternal" href="http://www.clips.ua.ac.be/pages/pattern-examples-100days"><img src="../g/pattern_example_100days.jpg" alt="" width="70" height="70" /><br /></a></p>
</td>
<td>&nbsp;</td>
<td><span class="smallcaps">web mining and visualization </span><span class="small">(2010)</span><br />This case study uses a number of different approaches to mine, correlate and visualize about 6,000 Google News items and 70,000 tweets.&nbsp;<br /><a href="http://www.clips.ua.ac.be/pages/pattern-examples-100days">read more »</a></td>
</tr>
</tbody>
</table>
</div>
</div></div>
</div>
</div>
</div>
</div>
</div>
</div>
<script>
SyntaxHighlighter.all();
</script>
</body>
</html>