An Approach of Content-based Recommend with Support Vector Regression Applied to E-Business

摘要

The recent advances of using recommender systems in e-business have arisen great interests around the world; many researches have surged to improve the recommend approaches. Traditionally there are three recommend approach genres: collaborative filtering approaches, content-based approaches and hybrid approaches. Support Vector Regression (SVR) algorithm has been introduced to construct a content-based recommend approach. First, the contents of rated items are analyzed with SVR to build regression model of user profiles for active users. Then the contents of unrated items are matched with user profiles. Based on the results corresponding recommendations are given out. Experimental results on the EachMovie dataset show that the proposed approach has better recommend performance and higher time efficiency than the conventional collaborative filtering approach.