/C-Cat

A collection of tools for applying word senses to large corpora

Primary LanguageJavaGNU General Public License v2.0GPL-2.0

The C-Cat library provides libraries for large scale text processing using the
hadoop framework.  It's ultimate goal is to provide tools and libraries for
automatically customizing a wordnet ontology based on the contents of a
particular corpus.  

It is structured into three sub-modules:
1) extendOntology-core: This provides a core set of text and collection like
classes.  
2) extendOntology-wordnet: This provides core wordnet libraries for reading,
writing, and modifying the wordnet hierarchy along with several synset
similarity metrics and several word sense disambiguation algorithms.
3) extendOntology: This is the complete package that includes both the core and
wordnet submodules.  On top of the two modules, it includes a text
pre-processing framework for documents stored in HBase.  This is the most
unstable module and is under heavy development.

This project utilized maven as it's build system.  Most of the library
dependencies are handled via maven, but a few jars are from libraries that have
not been mavenized yet.  

To install these jars into maven, run 

./add_non_maven_jars.sh

Then, build the entire project with

mvn package

This will create two jars in target: extendOntology-1.0.jar and
extendOntology-1.0-jar-with-dependencies.jar.  To run any of the mains provided
without maven, include both of these jars in the classpath.