A Hybrid Recommender System: Uniqueness of Choices by Using Machine Learning Technique

Abstract

The problem of recommending similar sets of items in the online business community is called item recommendation. An item recommendation aims to recommend a new item that matches the user’s interests. Universally, recommendation amenities have become significant due to their support in e-commerce applications like online shopping, digital promotions, and various research domains. The collaborative approach of the recommender engine filters out the k-nearest neighbours and then the similarity is compared between the neighbourhoods. In this paper, an algorithm is proposed, to recommend the items to the users with respect to the uniqueness of users’ choice. The result achieved is a mixture of both types of items, which are commonly and rarely bought by others. The proposed technique uses machine learning modules to learn actively and recommend.