An Adaptive Hesitant Fuzzy Sets Based Group Recommendation System

Abstract

Accurate group movie recommendation systems are a need in society today. We find that people tend to watch movies in groups rather than by themselves. However, the groups of people that tend to watch movies together are very diverse. In the existing methods, the characteristics of individual users are simply aggregated to obtain the group’s attributes and most of the time useful data is not utilized. This can be improved upon by ensuring the utilization of all the data that we are presented with from the scenario. The method proposed in this paper is termed integrated as we weighed in the individual traits of each user in the group when predicting the group’s rating for a movie. We used the concept of Hesitant Fuzzy Sets (HFS) in order to keep track of the characteristics of each of the users individually. The method we proposed in this paper employs Matrix Factorisation (MF) based Collaborative Filtering (CF) along with hesitant fuzzy sets. The way we performed MF based CF for a group is that we found the factors first and then formed the groups. The ratings were then predicted for these groups. The groups we have considered are of three sizes - 3 users, 5 users, and 10 users.

Systematic Selection of Blockchain Platforms Using Fuzzy Ahp-Topsis

“…In a fuzzy multiple-criteria decision-making (FMCDM) problem, the fuzzy set theory provides the advantage of easily using linguistic terms for alternative evaluation [21,41]. Many studies have compared different MCDM techniques based on their perspectives and theories [19,21,22].…”