Offering A Product Recommendation System in E-commerce

TL;DR

This paper presents a product recommendation system for e-commerce that combines explicit and implicit ratings, cosine similarity, and association rule mining to improve recommendation accuracy, especially for new users.

Contribution

It introduces a hybrid recommendation approach using weighted cosine similarity and association rules, considering transaction timing to enhance performance.

Findings

Implicit ratings yield acceptable recommendation performance.

Incorporating association rules improves system effectiveness.

Time-aware analysis helps address sequence recognition issues.

Abstract

This paper proposes a number of explicit and implicit ratings in product recommendation system for Business-to-customer e-commerce purposes. The system recommends the products to a new user. It depends on the purchase pattern of previous users whose purchase pattern is close to that of a user who asks for a recommendation. The system is based on weighted cosine similarity measure to find out the closest user profile among the profiles of all users in database. It also implements Association rule mining rule in recommending the products. Also, this product recommendation system takes into consideration the time of transaction of purchasing the items, thus eliminating sequence recognition problem. Experimental result shows for implicit rating, the proposed method gives acceptable performance in recommending the products. It also shows introduction of association rule improves the performance measure of recommendation system.