IURA: An improved user-based collaborative filtering method based on innovators

User based collaborative filtering (UserCF) is a method that generates recommendations based on the preferences and past actions of like-minded users. Currently, most UserCF based recommendation systems do not consider the users' purchase precedence and activeness when locating those like-minded users. Yet, these two factors contain valuable information that can contribute to recommendation accuracy and diversity. First, according to the Diffusion of Innovations Theory (DIT), the earlier that a like-minded user purchased an item, the more likely that he would be a trend leader in his respective area of interest. Such users are called the innovators, and they should have higher level of influence on their followers than a typical like-minded user. Second, innovators are typically more active and more adventurous users. They are the ones who are more willing to try out new products of various genres, and would therefore contribute the diversity of the recommendation. Based on these reasons, we propose in this paper an improved UserCF mechanism based on innovators instead of simply likeminded users. The proposed method is simple to implement, and also applicable even in the cases where item release time is not available. Extensive experiments were conducted to evaluate the proposed mechanism using various metrics and the results were encouraging: our proposed scheme not only achieved the best results in term of accuracy, but also performed well in terms of diversity (including intra-list and aggregate diversity) as well.