An Effective Deep Learning based Recommender System with user and item embedding

Abstract

The success of deep learning in a variety of fields has spurred a rush of desire to develop new novel recommender systems. As a result, our research introduces new RecDNNing, a ground-breaking technique that combines integrated people and products with a deep learning model. There are two phases to the suggested suggestion method. We begin by constructing a unique mathematical version for each human and item, which we refer to as user and item embedding. The embeddings of the goods and users then are aggregated, concatenated, and sent into the deep learning model. In the second phase, the spliced people and items implants are utilised as sources, and the ratings scores are forecasted using the forward propagation technique. The suggested new RecDNNing algorithm outperforms state-of-the-art techniques on MovieLens, according to the results.