A Comprehensive Review of CNN-based Recommendation Systems
Puneet Kumar Sharma, Chetan Bhardwaj, Dolly Sharma, Rishabh Mishra · 2024
In today’s world, customers have millions of choices of products on e-commerce websites. To help them in not getting overwhelmed by all this variety and make informed choices, these sites employ recommendation systems that provide their customers with suggestions on what to buy. Due to the existence of many techniques, the selection of a technique is a complex task. In addition, each of these have their own features, advantages, and disadvantages. The field of recommendation systems has seen many new groundbreaking and novel techniques from considering the content of items to considering the behavior of related users from implementing NLP techniques to deep learning models there is much research undertaken by the researchers; one of them is using CNNs for both NLP-based approaches or image-based approaches. This paper aims to undertake a systematic review on the literature related to the fields of recommender systems, with a focus on the domain of e-commerce and systems created using CNNs. In the first section of the paper an overview of these systems is given followed by a literature review consisting of comparison between various studies, there algorithms, evaluation metrics and their advantages. Finally, the paper concludes by drawing conclusion of the review and providing future scope for research in this field. In this way, our review paper provides a comprehensive view of the literature related to CNN-based recommender systems.