An Evaluation of Machine Learning Models for Sentiment Analysis: A Comparative Study
Maanav Thalapilly, G Kisor, Ketone Agasti, Madugula Vishnu Datta, Tricha Anjali, Ajith Vallat Nair · 2025
In this expansive research initiative, the central focus is on sentiment analysis, where the performance of four machine learning models across diverse datasets from Flipkart, Amazon, IMDb, and eBay is evaluated. Sentiment analysis, pivotal in shaping public opinion, is examined using Random Forest, Naïve Bayes, XGBoost, and Support Vector Classifier (SVC), providing insights into their adaptability and performance variations. Evaluation metrics such as accuracy, precision, and F1 score are meticulously examined to uncover nuanced insights. This research contributes to refining sentiment analysis methodologies and optimizing machine learning models for specific use cases, influencing decision-making processes, and enhancing customer satisfaction assessments. Practical implications extend to industries reliant on accurate sentiment interpretation, advancing academic understanding and informing future natural language processing endeavors.