Sentiment Based Product Recommendation System Using Machine Learning Techniques

Abstract

A sentiment-based product recommendation system is a system that utilizes natural language processing to extract sentiment features from reviews and the machine learning algorithms namely logistic regression, random forest classifier, XGBoost classifier and CatBoost classifier are applied to classify sentiments and these algorithms performance is evaluated based on the performance metrics such as accuracy, recall, precision and F1 score.Memory based collaborative filtering models such as user based and item-based recommendation models are developed and assessed using benchmark data, its performance is evaluated based on the evaluation metric Root mean square error (RMSE) and then the recommendations are fine tuned.While real-time data is obtained through web scraping from the Mamaearth website.Sentiment-based recommendations are implemented on real-time data, and their performance is evaluated using the same key performance metrics and the recommendations are fine-tuned.Finally, a Streamlit web app is created and deployed using Ngrok to enhance the accessibility and utility of the recommendation system.