TwitHub project

• an open-source school project that employs machine learning to classify tweets under the topic 'Harry Potter', but this is open for extensibility
• an executable demoing the results of modeling, validating and iterating can be in the form of TwitHub.jar

Dependencies

• The excellent Twitter4J
• Weka classification algorithms (SMO, Naive Bayes, J48) are being used here
• MorphAdorner's Lexicon, Part of Speech tagger, and Language Recognizer facilities
• Google spell check
• The dependencies of the above libraries

Some stuff…

• yes, I know, the code contains a real-life Twitter account, with the user name and password placed in clear text; along with a developer account CONSUMER KEY AND SECRET

I will be disabling these accounts soon… bring your own developer tokens :D

• This is an Eclipse project.

• feel free to poke around the code; it's still a little messy in com.twitminer.viewer.algorithms though.