Inferring User Interests from Relevance Feedback with High Similarity Sequence Data-Driven Clustering

Abstract

Relevance feedback is an important source of information about a user and often used for usage and user modeling for further personalization of user-system interactions. In this paper we present a method to infer the userpsilas interests from his/her relevance feedback using an online incremental clustering method. For inference of a new interest (concept) and concept update the method uses the similarity characteristics of uniform user relevance feedback. It is fast, easy to implement and gives reasonable clustering results. We evaluate the method against two different data sets, demonstrate and discuss the outcomes.