Learning the user profile in online music domain based on the semantic similarity

The endless amount of information on the web, known as "lost-in-hyper-space syndrome", easily overwhelms users. User profiles can be used as a means to support extracting relevant information by indicating user interests and filtering irrelevant information out.;There are two main issues in developing effective user profiles: first --- a cyclic approach that guarantees updating a user profile appropriately over time; second --- an effective representation of things that satisfies user interests. Machine learning methods can provide approaches to update a user profile, and Semantic Web technologies enable agents with a better understanding of contents on the web providing a higher level of services to the user. In this research work we propose a new method to develop and maintain a user profile by analyzing his/her access behavior on the web. We use Semantic Web techniques proposing a semantic similarity measure to ensure a better grasp of the user interests.