Personalizing the Search for Knowledge
Minko Dudev, Shady Elbassuoni, Julia Luxenburger, Maya Ramanath, Gerhard Weikum · 2008
Recent work on building semantic search engines has given rise to large graph-based knowledge repositories and facilities for querying them and more importantly, ranking the results. While the ranking provided may prove to be acceptable in general, for a truly satisfactory search experience, it is necessary to tailor the results according to the user’s interest. In this paper, we address the issue of personalizing query results in the specific setting of graph-based knowledge bases. In particular, we address two important issues: i) construction of the user profile based on the inference of the user’s interest and ii) a formal model for personalized scoring which incorporates the user’s interest. Preliminary experimental results show that our techniques are indeed promising.