A Genetic Approach for Enhancing Recommender System’s Stability

Abstract

With the growing scope of the E-commerce industry, a recommender system places a critical role in predicting correct entity to the user. Accuracy and degree of trust are most important parameters of a recommender system. Many recommender algorithms are proposed in the literature to enhance the accuracy of recommendations to the user. This paper is focused on enhancing the stability of recommender system using genetic and repetitive smoothening approach. Stability determines the degree of trust by the user to use the items recommended by a recommender system. SVD and slope one recommenders systems are used along with the proposed approach and it has been shown that stability is enhanced using the genetic-repetitive algorithm. Genetic approach has shown significant improvements in the field of research and is considered one of the robust techniques.