A recommendation system using machine learning techniques: a case study of oversized clothes in e-commerce platform

This research develops and compares the effectiveness of recommendation algorithms for a plus size clothing market on e-commerce platforms with the rapid growth of the e-commerce industry and the increasing importance of using recommendation systems to drive sales and customer satisfaction, the plus size clothing market also has significant growth. However, there is a lacking study on the recommendation system for a plus size cloth. The study compares the performance of association rule mining and collaborative filtering algorithms on a dataset of plus size clothing transaction on an e-commerce platform. The results of the models are evaluated using precision, recall and F1-score to identify the best-performing recommendation model for plus-size clothing. The User-Based (MSD) model emerges as the top performer with a precision rate of 29%, recall rate of 25%, and F1-score of 26%, outperforming User-Based (Pearson) and SVD models. On the other hand, FP-Growth exhibits the least performance with a precision rate of 7.88%, recall rate of 4.23%, and F1-score of 5%. The findings of this study have implications for e-commerce businesses and researchers in the field of recommendation systems. The results of this study contribute to a better understanding of the effectiveness of recommendation algorithms in the plus size clothing market.