Research on Semantic Natural Language Processing and Interfaces
Natural language processing and development of interfaces based on natural languages is important in many applications. When the content is restricted to a specific domain of knowledge better results can be obtained (less ambiguities arise) using domain ontolgies. In this line of research we developed methodologies for exploiting domain semantics in the natural language processing, and a software framework that facilitates the automatic generation of natural language interfaces to knowledge repositories. We are also pursuing research in the syntactic analysis of the Greek language.
In this scope TUC/MUSIC has developed systems that do automatic syntactic analysis and tested them in text corpora, as well as in a system for information retrieval using a natural language interface and text content with satisfactory results.
Stavros Christodoulakis Prof. stavros ced.tuc.gr