README

Processing Resource for GATE which uses DigitalPebble's TextClassification API (https://github.com/DigitalPebble/TextClassification/)

Compilation
Run 'ant distrib'. The GATE compile dependencies are managed by IVY, you need an internet connection so that it can fetch the GATE jars.

Installation
Unzip the distribution archive into GATE/plugins or to the directory of your choice then load it using GATE's plugin management panel

Usage
The plugins contains 3 Processing resources :
- TrainingCorpusCreator : generates a lexicon + raw file in the specified directory. See https://github.com/DigitalPebble/TextClassification/ for instructions on 
how to generate a vector file and model from a raw file.
- ClassifierPR : takes a model and lexicon to classify the annotations specified in textAnnotationType
- NGram maker : generates ngrams that can be used as input for the corpus generation or classification