Decision making operations: arithmetic mean, geometric mean, harmonic mean in user-based collaborative filtering of multi-criteria recommender systems

Abstract

The recommender system used to introduce the best products or services to users is always researched and improved by scientists. Today, users always ask for diversity in the search for the desired information. Therefore, in data stored, the recommender system should be considered on several criteria to make a rational decision to suggest to the user. One of the ways to do this is to build a multi-criteria recommender system that performs user-based collaborative filtering with a number of average operations that are used to make decisions. In this paper, we propose an new approach to choose efficient operations: Arithmetic mean (AM), Geometric mean (GM), Harmonic mean (HM) in user-based collaborative filtering of multi-criteria recommender systems. This model demonstrates the synergy and interplay between criteria (users) for decision making. The model was evaluated through experimentation with the multirecsys tool on two datasets: MovieLense100K and Jester5k. The experiment illustrated compare between the models each other. The results of the experiment have shown that we need to select the operation to apply in the system must adapt to the characteristics of each dataset. In such cases, the results of the consultancy will be highly effective