############################################# Verb Checker Framework for handling data in order to allow the testing of various automated verb error checking methods using Mallet classifiers Author: Noah Weber Email: webern2@winthrop.edu ############################################# Dependencies: -Mallet library is installed -Nltk along with the verbnet package is installed -FCE corpus is downloaded Handling FCE data: -Before running program concat all FCE xml files into one large xml file (and use sed or something to make this into valid xml). -Then split this data into testing and training sets (~0.9 training to testing proportion is fine now) -Use python process_fce_data.py fcexmlfile.xml trainout trainout_delim to get fce text data and delimited fce text data -Then use the annotate_text.sh script to pos tag both the fce text data file and the delimited fce data Use the prep_data script to run through data preperation pipeline Use run-classifier script to run the classifier and use eval_results.py to evaluate the results of the classifier Send questions and complaints to webern2@winthrop.edu
weberna/verb-checker
Various scripts and whatnot for automatic checking/correcting of verb related errors, backup of a local git repo
Python