Optimization and Machine Learning

5An Interactive Attention Network with Stacked Ensemble Machine Learning Models for Recommendations

Ahlem DRIF¹, SaadEddine SELMANI¹ and Hocine CHERIFI²

¹Computer Sciences Department, Ferhat Abbas University Setif 1, Algeria

²LIB, University of Burgundy Franche-Comté, Dijon, France

Recommender systems are broadly used to suggest goods (e.g. products, news services) that best match user needs and preferences. The main challenge comes from modeling the dependence between the various entities incorporating multifaceted information, such as user preferences, item attributes and users’ mutual influence, resulting in more complex features. To deal with this issue, we design a recommender system incorporating a collaborative filtering (CF) module and a stacking recommender module. We introduce an interactive attention mechanism to model the mutual influence relationship between aspect users and items. It allows the mapping of the original data to higher order feature interactions. In addition, the stacked recommender, composed of a set of regression models and a meta-learner, optimizes the weak learners’ performance with a strong learner. The developed stacking recommender considers the content for recommendation to create a profile model for each user. Experiments on real-world datasets demonstrate that the proposed algorithm can achieve more accurate predictions and higher recommendation efficiency.

5.1. Introduction

Recommender systems have become an integral part of e-commerce ...