A Machine Learning-Based Framework for Predicting E-Commerce Product Ratings

Big data has been the center of attention for industry due to the volume of data that is being generated and growing over time. Machine learning algorithms are being employed more and more to anticipate various features of products. Product ratings are one of the key metrics used by every firm to evaluate the performance of its products. The product rating methodology persuades customers to purchase any item. It also demonstrates a product's potential for success in the consumer market. In order to forecast a product's rating based on its category, listed price, reduced price, and number of ratings, this study used a machine learning model. The Amazon dataset, which is accessible on Kaggle, was used in this study. This study evaluates various machine learning models using a dataset of 61,950 electronics appliance entries from Amazon to predict product ratings based on category, listed price, reduced price, and number of ratings. The Random Forest model outperforms others, with a precision of 0.65 and recall of 0.6 for 1-star ratings, and demonstrates superior performance for 2-star, 4-star, and 5-star predictions.