Document Modeling Using Syntactic and Semantic Information

This paper proposes a novel document representation model that considers both semantic and syntactic information. The proposed model combines syntactic and semantic components using, respectively, syntactic dependency trees and a word sense disambiguation algorithm. The suitability of the proposed model has been demonstrated through an application of document clustering. The results show that the incorporation of semantic information in syntactic dependency trees improves the quality of the clustering solution compared with documents modeled by syntactic dependency trees only.