International Research Journal of Education and Technology

In product recommendation systems, traditional recommendation systems are often too focused on numeric ratings to extract Meaningful insights from user sentiments. Here, we propose a new concept: a user sentiment-based product recommendation system leverages the Valence Aware Dictionary and sentiment Reasoner (VADER) model to enhance Otherwise Obtaining accurate recommendations. The product recommendation system analyses user reviews by classifying user sentiment into a positive, negative, and neutral sentiment recommendations. Our proposed approach leverages Natural Language Processing (NLP) models and deep learning models that enhances both, recommendation accuracy and user satisfaction while utilizing ongoing user opinions to enhance performance. Our experimental analysis indicates robust performance by the model on classifying sentiment and recommendation accuracy making it a valuable asset for ecommerce retailers and businesses.

Index Terms:

            Sentiment Analysis, Product Recommendation, VADER, Natural Language Processing, User Experience, Machine Learning.