%0 Journal Article %1 keyhere %A Degemmis, Marco %A Lops, Pasquale %A Semeraro, Giovanni %D 2007 %J User Modeling and User-Adapted Interaction %K collaborative filtering recommendersystems wordnet %N 3 %P 217--255 %T A content-collaborative recommender that exploits WordNet-based user profiles for neighborhood formation %U http://dx.doi.org/10.1007/s11257-006-9023-4 %V 17 %X Abstract  Collaborative and content-based filtering are the recommendation techniques most widely adopted to date. Traditional collaborative approaches compute a similarity value between the current user and each other user by taking into account their rating style,that is the set of ratings given on the same items. Based on the ratings of the most similar users, commonly referred to as neighbors, collaborative algorithms compute recommendations for the current user. The problem with this approach is that the similarityvalue is only computable if users have common rated items. The main contribution of this work is a possible solution to overcomethis limitation. We propose a new content-collaborative hybrid recommender which computes similarities between users relyingon their content-based profiles, in which user preferences are stored, instead of comparing their rating styles. In more detail,user profiles are clustered to discover current user neighbors. Content-based user profiles play a key role in the proposedhybrid recommender. Traditional keyword-based approaches to user profiling are unable to capture the semantics of user interests.A distinctive feature of our work is the integration of linguistic knowledge in the process of learning semantic user profiles representing user interests in a more effective way, compared to classical keyword-based profiles, due to asense-based indexing. Semantic profiles are obtained by integrating machine learning algorithms for text categorization, namelya nave Bayes approach and a relevance feedback method, with a word sense disambiguation strategy based exclusively on thelexical knowledge stored in the WordNet lexical database. Experiments carried out on a content-based extension of the EachMoviedataset show an improvement of the accuracy of sense-based profiles with respect to keyword-based ones, when coping with thetask of classifying movies as interesting (or not) for the current user. An experimental session has been also performed inorder to evaluate the proposed hybrid recommender system. The results highlight the improvement in the predictive accuracyof collaborative recommendations obtained by selecting like-minded users according to user profiles.

@article{keyhere, abstract = {Abstract  Collaborative and content-based filtering are the recommendation techniques most widely adopted to date. Traditional collaborative approaches compute a similarity value between the current user and each other user by taking into account their rating style,that is the set of ratings given on the same items. Based on the ratings of the most similar users, commonly referred to as neighbors, collaborative algorithms compute recommendations for the current user. The problem with this approach is that the similarityvalue is only computable if users have common rated items. The main contribution of this work is a possible solution to overcomethis limitation. We propose a new content-collaborative hybrid recommender which computes similarities between users relyingon their content-based profiles, in which user preferences are stored, instead of comparing their rating styles. In more detail,user profiles are clustered to discover current user neighbors. Content-based user profiles play a key role in the proposedhybrid recommender. Traditional keyword-based approaches to user profiling are unable to capture the semantics of user interests.A distinctive feature of our work is the integration of linguistic knowledge in the process of learning semantic user profiles representing user interests in a more effective way, compared to classical keyword-based profiles, due to asense-based indexing. Semantic profiles are obtained by integrating machine learning algorithms for text categorization, namelya nave Bayes approach and a relevance feedback method, with a word sense disambiguation strategy based exclusively on thelexical knowledge stored in the WordNet lexical database. Experiments carried out on a content-based extension of the EachMoviedataset show an improvement of the accuracy of sense-based profiles with respect to keyword-based ones, when coping with thetask of classifying movies as interesting (or not) for the current user. An experimental session has been also performed inorder to evaluate the proposed hybrid recommender system. The results highlight the improvement in the predictive accuracyof collaborative recommendations obtained by selecting like-minded users according to user profiles.}, added-at = {2007-10-14T10:54:16.000+0200}, author = {Degemmis, Marco and Lops, Pasquale and Semeraro, Giovanni}, biburl = {https://www.bibsonomy.org/bibtex/2a3dbdb2ec174dd4a7b8f88c02d8eb1da/wnpxrz}, description = {A content-collaborative recommender that exploits WordNet-based user profiles for neighborhood formation}, interhash = {2f3a2687637a2b0d3c6ac329f9342b9e}, intrahash = {a3dbdb2ec174dd4a7b8f88c02d8eb1da}, journal = {User Modeling and User-Adapted Interaction}, keywords = {collaborative filtering recommendersystems wordnet}, month = {#jul#}, number = 3, pages = {217--255}, timestamp = {2007-10-14T10:54:16.000+0200}, title = {A content-collaborative recommender that exploits WordNet-based user profiles for neighborhood formation}, url = {http://dx.doi.org/10.1007/s11257-006-9023-4}, volume = 17, year = 2007 }