Evaluation of user-based algorithms used in the recommendation system

Abstract

Nowadays, the consumers’ demand for online shopping is rapidly increasing. To satisfy the users’ satisfaction, service providers have come up with many solutions to support the users in searching for the best items. In this paper, we examine a number of user-based algorithms in collaborative filtering for user recommendations, which is based on the previous users’ behaviors. The experiment was performed on the two data sets called “MovieLens” and “EachMovie”. The results showed that the Euclidean algorithm produces the best results. This algorithm might be used in online trading systems to improve the searching efficiency.

Keywords: User-based algorithms, quality of service, recommendation system.