Developing Natural Language Processing-based Methods for Text Classification

Funding Agency: Bogazici University Research Fund, BAP (Project 05A103D)

Project Manager: Tunga Güngör

Dates: 2005-2008

In this project, natural language processing techniques were used for the text classification problem. Currently, many research studies are conducted for text classification purposes and some of them have practical applications. However, the success rate of these studies has not exceeded a certain limit.

The main reason is that almost all studies use only syntactic information and do not utilize semantic information. In other words, words in texts are handled independently of their meanings. In this project, new methods were proposed to address this deficiency and to classify texts by taking into account the meanings they contain.