OxiNet: A New Natural Language Processing Software

Report Number:  TR-GWU-CS-2009-005
Submit Date:  2009-12-04
Author:  Ganapathy Mani, Stephen Kaisler

The idea of giving computers the ability to process human language is as old as the idea of computer themselves. Nowadays, designing computer systems with full human intelligence capabilities has been the conspicuous focus of Artificial Intelligence researchers. They have made tremendous amount of efforts in designing programs that understand with deep knowledge of the semantic meaning of texts. In this paper, I am going to introduce an art called Natural Language Processing that makes all the Artificial Intelligence researchers to reach their goal. Through this paper, I will introduce the most important Natural Language Processing typical examples (UNO, Rule based, Statistical, Finite state machine, Logic programming etc.). Moreover, I will give a brief overview on the important tools used in the implementation of NLP applications. These tools include Lexicons such as Wordnet and syntactic parsers such as the Link Grammar. Finally I will provide a design schema that illustrates how lexicons and parsers are used in order to implement a simple NLP application. The application will take a user-defined sentence as input and it detects the main verb and outputs all the possible synonyms corresponding to that verb. This work presents a new Wordnet model called, Optimized Lexicon Wordnet (OxiNet) that is able to detect anomalies in sentences and words of a given text with close to 80% accuracy.