README

		>> Twitter Sentiment Classifier <<

A simple Maven project that builds a sentiment classifier that is used to predict tweet sentiments.

Extracted features are stored in Redis using the jedis api and classification is done using Weka.

Training done using twitter sentiment corpus available at http://twittersentiment.appspot.com

Eclipse 'maven integration plugin' used for dependency management and 'checkstyle' for auto code formatting.

JUnit tests to check if required resources are in place (Stanford Part-of-Speech tagger and Redis Connection).


Build :
Script in pom.xml to build jar-with-dependencies (automated in the case of build servers, e.g. Jenkins),
else, if to be done manually, navigate to root of project and execute `mvn package`, the jar will appear
in `target` directory, under root, on successful build.

Note :
if maven plugin for eclipse does not download a dependency automatically, manually install artifact by,
e.g `mvn install:install-file -Dfile=commons-pool-1.6.jar -DgroupId=commons-pool -DartifactId=commons-pool
     -Dversion=1.6 -Dpackaging=jar -DgeneratePom=true`