We have experience on parsers, POS taggers and Named Entity recognizers on different open source tools like Stanford NLP, OpenNLP, Chaniak and Lingpipe. Wordnet has been an excellent tool and quite instrumental in implementing techniques like synonym clubbing. We have also implemented algorithms on Pronomial core reference resolution and word sense disambiguation as well as solved problems using HMM (Hidden Markov Model).

Our basic work has been focussed on analyzing CGM (Consumer Generated Media) like blogs, forums and reviews. Our work involves crawling websites, storing the content in the database and then automatically clustering and categorizing the content (Weka Toolkit).