-
Notifications
You must be signed in to change notification settings - Fork 5
Fork of WikipediaMiner Toolkit 1.1 to apply fixes
License
orph/wikipedia-miner
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<html>
<head>
<title>Wikipedia Miner - Readme</title>
<style>
body {
font-size: 14px ;
}
.code {
font-family: monospace;
font-weight: bold;
color: rgb(50, 50, 50) ;
}
.small{
font-weight: normal;
font-size: 13px ;
margin-top: 0px ;
margin-bottom: 4px ;
}
.bold {
font-weight: bold ;
}
</style>
</head>
<body>
<h1>Wikipedia Miner Readme</h1>
<p>
</p>
<ul>
<li><a href="#requirements">Requirements</a></li>
<li><a href="#installation">Installation</a></li>
<li><a href="#web">Web Services</a></li>
<li><a href="#data">Tables and Summary Files</a></li>
<li><a href="#licence">Licence</a></li>
</ul>
<h2><a name="requirements"></a>Requirements</h2>
<p>
To run wikipedia miner, you will need lots of hard-drive space and around
3G of memory. On top of that, you will need...
</p>
<ul>
<li>Write access to a <a href="http://www.mysql.com/">MySql</a> Server</li>
<li>Java (1.5 or above)</li>
<li><a href="http://dev.mysql.com/downloads/connector/j/">MySql Connector/J</a>—Java API for connecting to MySql databases.</li>
<li><a href="http://trove4j.sourceforge.net/">Trove</a>—Java API for efficient sets, hashtables, etc.</li>
<li><a href="http://www.cs.waikato.ac.nz/ml/weka">Weka</a>—an open-source workbench of machine learning algorithms for data mining tasks. </li>
</ul>
<p>
If you only need Wikipedia's structure rather than it's full textual content, then you can save a lot of time by using one of our pre-summarized
dumps (available <a href="">here</a>). Otherwise, you will need:
</p>
<ul>
<li>
A copy of Wikipedia's content (one of the <i>pages-articles.xml.bz2</i> files from <a href="http://download.wikimedia.org/backup-index.html">here</a>)
</li>
<li><a href="http://www.perl.org/get.html">Perl</a></li>
<li><a href="http://search.cpan.org/~triddle/Parse-MediaWikiDump-0.51/">Parse:MediaWikiDump</a>—Perl tools for processing MediaWiki dump files.</li>
</ul>
<p>If you want to host your own Wikipedia Miner web services, then you will also need</p>
<ul>
<li><a href="http://tomcat.apache.org/">Apache Tomcat</a></li>
</ul>
<h2><a name="installation"></a>Installation</h2>
<ol>
<li class="bold">
Set up a mysql server
<p class="small">
This should be configured for largish MyISAM databases. You will need a database to store the wikipedia data, and an account that has write access to it.
</p>
</li>
<p class="small">
If you are happy using the versions of Wikipedia that we have preprocessed and made available, then <a href="https://sourceforge.net/project/showfiles.php?group_id=191683&package_id=300169">download one from SourceForge</a>, and skip to <a href="#step5">step 5</a>. Otherwise...
</p>
<li class="bold">
Download and uncompress a <a href="http://download.wikimedia.org/enwiki/">wikipedia xml dump</a>
<p class="small">We want the file with current versions of article content, the one that ends with <i>pages-articles.xml</i>. Uncompress it and put it in a folder on its own, on a drive where you have lots of space.</p>
</li>
<li class="bold">
Tweak the extraction script for your version of Wikipedia
<p class="small">
This is only necessary if not working with the <i>en</i> (English) Wikipedia. Open up <span class="code">extraction/extractWikipediaData.pl</span> in a text editor. Under the section called <i>tweaking for different versions of wikipedia</i>, modify the following variables:
<ul class="small">
<li>
<span class="code">@disambig_templates</span>—an array of template names that are used to identify disambiguation pages
</li>
<li>
<span class="code">@disambig_categories</span>—an array of category names that disambiguation pages belong to. This is only needed because a lot of people do this directly, instead of using the appropriate template
</li>
<li>
<span class="code">$root_category</span>—the name of the root category, the one from which all other categories descend.
</li>
</ul>
</p>
<p class="small">
<em>WARNING:</em> These scripts have only been tested on <i>en</i> and <i>simple</i> Wikipedias. If you get a language version working, then please post the details up on the sourceforge forum.
</p>
</li>
<li class="bold">
Extract csv summaries from the xml dump
<p class="small">
The perl scripts for this are in the <span class="code">extraction</span> directory.
</p>
<ol style="list-style:lower-alpha">
<li class="bold">
<p class="small">
The main script, extractWikipediaData, does everything that most users will need (more on what it doesn't do in 4b). To run it, call
<span class="code">perl extractWikipediaData <dir></span>
where <span class="code">dir</span> is the directory where you put the xml dump.
</p>
<p class="small">
You can supply an extra flag <span class="code">-noContent</span> if you just want the structure of how pages link to each other rather than
their full textual content. This takes up a huge amount of space, and you can do a lot (finding topics, navigating links,
identifying how topics relate to each other, etc) without it.
</p>
<p class="small">
<em>WARNING:</em> this will keep the computer busy for a day or two; you can <a href="https://sourceforge.net/forum/forum.php?thread_id=2643967&forum_id=894561">check out the forums</a> to see how long it has taken other people.
If you do need to halt the process then don't worry, it will pick up more-or-less where it left off.
</p>
<p class="small">
If the script seems to stall, particularly when summarizing anchors or the links in to each page, then chances are you have run out
of memory. There is another flag <span class="code">-passes</span> to split the data up and make it fit. The default is <span class="code">-passes 2</span>; try something higher and
see how it goes.
</p>
</li>
<li class="bold">
<p class="small">
The only csv file that the above script will not extract is <span class="code">anchor_occurance.csv</span>, which compares how often anchor terms are used as links,
and how often they occur in plain text. Chances are you won't want this—it's mainly useful for identifying how likely terms are
to correspond to topics, so that topics can be recognized when they occur in plain text.
</p>
<p class="small">
This takes a very long time to calculate. Longer than extracting all of the other files. Fortunately it is easily parallelized.
The following scripts are included so that you can throw multiple computers (or processors) at the problem, if you have them.
<ul class="small">
<li>
<span class="code">splitData.pl <dir> <n></span>
<p style="margin-left: 10px">
splits the data found in <span class="code">dir</span> into <span class="code">n</span> seperate files.
The files are saved as <span class="code">split_0.csv</span>, <span class="code">split_1.csv</span>, etc. within the provided directory.
</p>
</li>
<li><span class="code">extractAnchorOccurances.pl </span>
<p style="margin-left: 10px">
calculates anchor occurrences from a split file. The directory must contain one of the split files produced by <span class="code">splitData.pl</span>,
and the <span class="code">anchor.csv</span> file created by <span class="code">extractWikipediaData.pl</span>. The results are saved to <span class="code">anchor_occurance_<n>.csv</span>.
</p>
</li>
<li><span class="code">mergeAnchorOccurances.pl <dir> <n></span>
<p style="margin-left: 10px">
merges all of the intermediate results. The <span class="code">dir</span> must contain all of the seperate <span class="code">anchor_occurance_<n></span>.csv files.
The result is saved to <span class="code"><dir>/anchor_occurance.csv</span>
</p>
</li>
</ul>
</p>
</li>
</ol>
</li>
<li class="bold">
<a name="step5"></a>Import the extracted data into MySQL
<p class="small">
The easiest way to do this is via java—just create an instance of WikipediaDatabase with the details of the database you created
earlier, and call the loadData() method with the directory containing the extracted csv files.
This will do the work of creating all of the tables and loading the data into them, and will even give you information on
how long it's taking. At worst this should take a few hours.
</p>
<p class="small">
The details of this are in the JavaDoc.
</p>
<p class="small">
<em>NOTE:</em> You need the MySQLConnectorJ, Trove, and Wikipedia-Miner jar files in the build path to compile and run the java code.
You may also need to increase the memory available to the Java virtual machine, with the -Xmx flag
</p>
</li>
<li class="bold">
Delete unneeded files
<p class="small">
Don't delete everything. Some of the csv files will be needed for caching data to memory, because its faster to do
that from file than from the database. So keep the following files:
<ul class="small">
<li class="code">page.csv</li>
<li class="code">categorylink.csv</li>
<li class="code">pagelink_out.csv</li>
<li class="code">pagelink_in.csv</li>
<li class="code">anchor_summary.csv</li>
<li class="code">anchor_occurance.csv</li>
<li class="code">generality.csv</li>
</ul>
</p>
<p class="small">
You can delete all of the others. It might be worth zipping the original xml dump up and keeping that though, because they
don't seem to be archived anywhere for more than a few months.
</p>
</li>
<li class="bold">
Start developing!
<p class="small">
Hopefully the JavaDoc will be clear enough to get you going. Also have a look at the main methods for
each of the main classes (Wikipedia, Article, Anchor, Category, etc) for demos on how to use them.
</p>
<p class="small">
Pop into the <a href="https://sourceforge.net/forum/?group_id=191683">SourceForge forum</a> if you have any trouble.
</p>
</li>
</ol>
<h2><a name="web"></a>Web Services</h2>
<p>
If you need help on using any of the <a href="http://wdm.cs.waikato.ac.nz:8080">Wikipedia Miner web services</a>, go <a href="http://wdm.cs.waikato.ac.nz:8080/service?help">here</a> or append <span class=code>&help</span> to any of the URLs.
</p>
<p>
If you want to host the <a href="http://wdm.cs.waikato.ac.nz:8080">Wikipedia Miner web services</a> yourself, you will need to:
<ol>
<li class=bold>
Get Apache Tomcat up and running
<p class=small>This is beyond the scope of this document, but there are lots of tutorials out there.</p>
</li>
<li class=bold>
Ensure that the Wikipedia Miner web directory is accessable via Tomcat
<p class=small>This is the <span class=code>web</span> directory within your Wikipedia Miner installation. Again, there are lots of tutorials out there to explain how to do this.</p>
</li>
<li class=bold>
Gather the appropriate jar files
<p class=small>Place the jar files for Wikipedia Miner, Weka, Trove and MySql Connector/J into <span class=code>web/WEB-INF/lib/</span>.</p>
</li>
<li class=bold>
Configure web.xml
<p class=small>Edit <span class=code>web/WEB-INF/web.xml</span> to specify the following context parameters:</p>
<ul class=small>
<li><span class=code>server_path</span>
<p class=small style="margin-left:10px">
The url of the <span class=code>web</span> directory (as it would be typed into a web browser)
</p>
</li>
<li><span class=code>service_name</span>
<p class=small style="margin-left:10px">
The name of the service (typically <i>service</i>, unless you change it)
</p>
</li>
<li><span class=code>mysql_server</span>
<p class=small style="margin-left:10px">
The name of the server hosting the mysql database (typically <i>localhost</i>)
</p>
</li>
<li><span class=code>mysql_database</span>
<p class=small style="margin-left:10px">
The name of the database in which Wikipedia Miner's data has been stored.
</p>
</li>
<li><span class=code>mysql_user</span>
<p class=small style="margin-left:10px">
The name of a mysql user who has read access to the database (you can leave this out if anonymous access is allowed)
</p>
</li>
<li><span class=code>mysql_password</span>
<p class=small style="margin-left:10px">
The password for the mysql user (you can leave this out if anonymous access is allowed)
</p>
</li>
<li><span class=code>data_directory</span>
<p class=small style="margin-left:10px">
The directory containing csv files extracted from a Wikipedia dump.
</p>
</li>
</ul>
<p class=small>
If you want users to be able to wikify URLs (the <i>wikify</i> service) or retrieve images from FreeBase (the <i>define</i> service) then you
will have to tell Wikipedia Miner how to connect to the internet.
</p>
<ul class=small>
<li><span class=code>proxy_host</span></li>
<li><span class=code>proxy_port</span></li>
<li><span class=code>proxy_user</span></li>
<li><span class=code>proxy_password</span></li>
</ul>
<p class=small>
If you want users to be able to wikify anything, then you will have to tell Wikipedia miner where to find models for disambiguation and link detection.
</p>
<ul class=small>
<li><span class=code>wikifier_disambigModel</span>
<p class=small style="margin-left:10px">
You can create one of these by building and saving a Disambiguator classifier (see the JavaDoc), or just use the one provided in <span class=code>models/disambig.model</span>.
</p>
</li>
<li><span class=code>wikifier_linkModel</span>
<p class=small style="margin-left:10px">
You can create one of these by building and saving a LinkDetector classifier (see the JavaDoc), or just use the one provided in <span class=code>models/linkDetect.model</span>.
</p>
</li>
<li><span class=code>stopword_file</span>
<p class=small style="margin-left:10px">
A file containing words to be ignored when detecting links (one word per line).
</p>
</li>
</ul>
</li>
</ol>
</p>
<p>
With Tomcat running, you should now be able to navigate to <span class=code>server_path</span> in a web browser and see a page like <a href="http://wdm.cs.waikato.ac.nz:8080">this</a>, listing all of the web services available.
</p>
<h2><a name="data"></a>Tables and Summary Files</h2>
<p>This section describes the summaries (csv files and database tables) that Wikipedia miner produces, in case you want to use them directly. </p>
<ul>
<li class="bold">
page
<p class=small>lists all of the valid pages in the dump, along with their titles and types</p>
<ul class=small>
<li>id (Integer)</li>
<li>title (String)</li>
<li>type (1=ARTICLE, 2=CATEGORY, 3=REDIRECT, 4=DISAMBIGUATION)</li>
</ul>
</li>
<li class="bold">
redirect
<p class=small>associates all redirect pages with their targets</p>
<ul class=small>
<li>from_id (Integer)</li>
<li>to_id (Integer)</li>
</ul>
</li>
<li class="bold">
categorylink
<p class=small>associates all categories with their child categories and articles.</p>
<ul class=small>
<li>parent_id (Integer)</li>
<li>child_id (Integer)</li>
</ul>
</li>
<li class="bold">
pagelink
<p class=small>associates all articles with the other articles they link to. Redirects are resolved wherever possible.</p>
<ul class=small>
<li>from_id (Integer)</li>
<li>to_id (Integer)</li>
</ul>
</li>
<li class="bold">
anchor
<p class=small>associates the text used within links to the pages that they link to.</p>
<ul class=small>
<li>text (String, the anchor text found within the link)</li>
<li>to_id (Integer)</li>
<li>count (Integer, number of times this anchor is used to link to this destination)</li>
</ul>
</li>
<li class="bold">
disambiguation
<p class=small>associates disambiguation pages with the senses listed on them.</p>
<ul class=small>
<li>from_id (Integer, the id of the disambiguation page)</li>
<li>to_id (Integer, the id of the sense page)</li>
<li>index (Integer, the position or index of this sense. Items mentioned earlier tend to be more obvious or well known senses of the term)</li>
<li>scope_note (String, the text used to explain how this sense relates to the ambiguous term)</li>
</ul>
</li>
<li class="bold">
equivalence
<p class=small>associates categories and articles which correspond to the same concept</p>
<ul class=small>
<li>cat_id (Integer)</li>
<li>art_id (Integer)</li>
</ul>
</li>
<li class="bold">
generality
<p class=small>lists the minimum distance (or depth) between a category or article and the root of Wikipedia's category structure. Small values indicate general topics, large values indicate highly specific ones.</p>
<ul class=small>
<li>id (Integer)</li>
<li>depth (Integer)</li>
</ul>
</li>
<li class="bold">
translation
<p class=small>associates an article with all of the language links listed within it.</p>
<ul class=small>
<li>id (Integer)</li>
<li>lang (String, <i>en</i>, <i>fr</i>, <i>jp</i>, etc)</li>
<li>text (String, generally a translation of the article's title, in the given language)</li>
</ul>
</li>
<li class="bold">
content
<p class=small>associates an article with it's content, in the original markup</p>
<ul class=small>
<li>id (Integer)</li>
<li>content (String, in raw mediawiki markup)</li>
</ul>
</li>
<li class="bold">
stats
<p class=small>provides some summary statistics of the wikipedia dump</p>
<ul class=small>
<li>article_count (Integer)</li>
<li>category_count (Integer)</li>
<li>redirect_count (Integer)</li>
<li>disambiguation_count (Integer)</li>
</ul>
</li>
<li class="bold">
anchor_occurance
<p class=small>lists the number of articles in which an anchor occurs as a link, vs the number of articles it occurs in at all.</p>
<ul class=small>
<li>anchorText (String)</li>
<li>link_count (Integer)</li>
<li>occ_count (Integer)</li>
</ul>
</li>
</ul>
<h2><a name="licence"></a>Licence</h2>
<p>
The Wikipedia Miner toolkit is open-source software, distributed under the terms of the <a href="http://www.gnu.org/copyleft/gpl.html">GNU General Public License</a>.
It comes with ABSOLUTELY NO WARRANTY.
</p>
</body>
</html>
About
Fork of WikipediaMiner Toolkit 1.1 to apply fixes
Resources
License
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published