Enhancing Recommendation Quality of Content-based Filtering through Collaborative Predictions and Fuzzy Similarity Measures

Abstract

Recommender systems (RSs) provide personalized suggestions about items to users while interacting with the large spaces on the web. Content based recommender systems (CB-RSs) offer personalized recommendations to a user mainly based on his past history and representations of the items. Although CB-RSs have been applied successfully in various domains, however recommendation diversity, representation of items as well as users’ modeling are still major concerns. Our work in this paper is an attempt towards developing effective content based filtering (CBF) by introducing an item representation scheme, fuzzy similarity measures and incorporating collaborative diverse predictions for alleviating its recommendation diversity. Experimental results show that the proposed hybrid scheme Fuzzy-CF-CBF outperforms hybrid CF-CBF, as well as both the fuzzy collaborative filtering (Fuzzy-CF) and the fuzzy content based filtering (Fuzzy-CBF).