Enhancing Product Review System: Leveraging CUDA-BB Algorithm for Sentiment Classification

Sentiment analysis is an important method in natural language processing (NLP) for detecting emotions and categorizing text as positive, negative, or neutral. It is invaluable for evaluating customer feedback, which is becoming increasingly accessible due to the massive amounts of data published on sites like as Facebook, Twitter, Amazon, and Flipkart. Businesses rely heavily on feedback to improve product quality and maintain a competitive advantage. To improve sentiment analysis, the CUDABB technique is proposed, which uses GPU-based parallel computing to forecast overall smart-phones star ratings. This strategy reduces bias and variation, which improves forecast accuracy and operating efficiency. The approach excels in forecasting smart-phones ratings and demonstrates the value of GPU computing (parallel processing) in sentiment analysis. With a $95.5 \%$ accuracy rate, the proposed classifier outperformed other base models and provided insightful customer feedback to support corporate decision-making.