From Logistic Regression to BERT: Benchmarking Sentiment Analysis Models on E-commerce Data

Product reviews play a crucial role in influencing consumer purchasing decisions and shaping brand perceptions. They provide valuable insights into product strengths and weaknesses, helping potential buyers make informed choices. However, analyzing and extracting meaningful insights from large volumes of review data can be challenging. AI-based techniques such as sentiment analysis have emerged as powerful tools for managing extensive review data. Sentiment analysis enables businesses to categorize reviews as positive, negative, or neutral, providing a comprehensive view of customer sentiment towards their products. By leveraging these AI techniques, businesses can gain deeper insights into customer preferences, identify emerging trends, and address issues highlighted in reviews. This paper analyzes the performance of legacy techniques like random forest and logistic regression, as well as AI techniques like the BERT model for sentiment analysis. Additionally, the project introduces Sentiscope, an application designed to analyze individual reviews, helping businesses make informed decisions to enhance their products and customer satisfaction.