Clustering to Deal with the New User Problem

Collaborative filtering (CF) techniques attempt to alleviate information overload by identifying which items a user will find interesting to browse. It focuses on identification of other users with similar tastes and usage of their opinions in order to recommend items. Commonly, however, CF suffers from the so-called new user problem which occurs when a new user is added to the system and there is not enough information to make a good suggestion. The system has to acquire some data about the new user in order to start making personalized recommendations. In this paper, we present a novel algorithm that combines previously acquired knowledge from article and user clustering in order to quickly determine the new user's interests. We attempt to address the new user problem by providing a personalized strategy for prompting the user with articles to rate. Our approach makes use of hypernyms extracted from the WordNet database and proves to be converging fast to the actual user interests based on minimal user ratings which are provided during the registration process.