Assessment of recommendation modelling with machine learning

The proliferation of limitless content brought about by the digital age has resulted in an extreme choice paradigm. It is possible for a user who is new to the site or searching for a species to become lost in this vast expanse. As a result, it’s essential to create a system that can direct consumers based on their interests. The Recommendation System (RS) was developed as a solution to this issue. RS is a tool that suggests different products according to user’s liking. Given the ability to address numerous issues related to over-selection in numerous web applications and widespread use, the advantages of the RS cannot be overstated. Machine learning (ML) has garnered a lot of attention recently in a variety of research fields, including natural language processing (NLP) and pattern recognition etc. This is due in part to ML’s remarkable performance as well as its alluring ability to demonstrate learning from scratch. When ML approaches are used to prediction and recommender systems, their impact becomes evident. In order to help future researchers in the field of recommender systems develop an effective system, this paper will conduct a systematic review of various recent contributions made in the field, with a focus on diverse applications such as movies, books, products, etc. We will analyse case studies conducted over the previous thirteen years (2011–2023), with a focus on 58 key studies selected from 2580 research papers found using the CADIMA tool. The goal of this systematic literature review (SLR) is to identify the best evidence-based strategy by evaluating several recommender system approaches applied to diverse applications. This review concludes by offering a much-needed summary of the state of the art in this area and highlighting the gaps and difficulties that still need to be filled.