International Journal of Computer Science and Applications c  Technomathematics Research Foundation Vol. 7 No. 1, pp. 1 - 22, 2010 ONTOLOGY BASED USER MODELING FOR PERSONALIZED INFORMATION ACCESS PLABAN KUMAR BHOWMICK ∗ Computer Science & Engineering, Indian Institute of Technology, Kharagpur West Bengal, India-721302 plaban@gmail.com SUDESHNA SARKAR Computer Science & Engineering, Indian Institute of Technology, Kharagpur West Bengal, India-721302 sudeshna@cse.iitkgp.ernet.in ANUPAM BASU Computer Science & Engineering, Indian Institute of Technology, Kharagpur West Bengal, India-721302 anupam@iitkgp.ac.in User modeling is an integral part of any personalized information retrieval system. The user model should be adaptable in order to capture the change in information needs of the users. In this paper, we present an ontology based user modeling strategy in the context of personalized information access. We have adopted a hybrid approach by capitalizing on the features of static and dynamic user proﬁling strategies. Static user proﬁle speciﬁes the user’s interest in a very focused manner and dynamic user proﬁling adds the feature of adaptability into it. The dynamic user proﬁling strategy make use of the data sources like usage log and mouse operations that are performed by the users during the browsing sessions. Experiments have performed to evaluate the proposed method for user proﬁling. Keywords : Ontology; user modeling; personalization; user proﬁle; usage log. 1. Introduction Retrieving personalized information from the ever increasing information space of Internet is a broad area of research. Due to the sheer volume of the web, searching for speciﬁc information from it has become tedious and time consuming. To alleviate this problem, eﬀorts have been started to customize the view of the web in a user speciﬁc way. The ﬁeld of Personalized Information Retrieval deals with the idea of retrieving speciﬁc information customized to the need of the user. The model of user is an integral part of any system that provides personalized information. User ∗ corresponding author 1