Evaluating the performance of a neural network-based multi-criteria recommender system

Abstract

Frequent use of internet applications and rapid increase in volumes of resources have made it difficult for online users to effectively make decisions on the kinds of information or items to select. Recommender systems (RSs) are intelligent decision-support tools that exploit preferences of users and suggest items that might be interesting to them. RSs are one of the various solutions proposed to address the problems of information overload. Traditionally, RSs use single rating techniques to predict and represent preferences of users for items that are not yet seen. Multi-criteria RSs use multiple ratings to various attributes of items for improving prediction and recommendation accuracy of the systems. However, one major challenge of multi-criteria RSs is the choice of an efficient approach for modelling the criteria ratings. Therefore, this paper aimed at employing artificial neural networks to model the criteria ratings and determine the predictive performance of the systems based on aggregation function approach. Seven evaluation metrics have been used to evaluate and measure the accuracy of the systems. The empirical results of the study have shown that the proposed technique has the highest prediction and recommendation than the corresponding traditional technique.

A Multi-Criteria Recommendation Framework using Adaptive Linear Neuron

“…For instance, one may think that u 1 and u 2 are similar since they all give the same ratings to i 1 , i 2 , and i 3 , and also they have almost the same opinions on the last item. But on the other hand, looking critically at their criteria ratings, the two users have entirely contradicting opinions; because, unlike u 2 , u 1 did not care about the influence of the rating given to the second criteria of each item [27]. Moreover, since MCRSs recommendation problems require multiple ratings, then the utility function presented in Eq.…”