Sentimental Analysis for Software Engineering
Ullal Akshatha Nayak, Vyshnavi R Hegde, P Sanchitha, Arun G S, Nikhil M Kulkarni · 2025
The rapid growth of online shopping underscores the importance of understanding customer behavior through user feedback. This paper presents a web application designed to perform real-time sentiment analysis on e-commerce platforms. Leveraging natural language processing and machine learning algorithms such as Support Vector Machines (SVM), Naive Bayes, Logistic Regression, and advanced models like Bidirectional Encoder Representations from Transformers (BERT), the system categorizes user reviews into positive, negative, or neutral sentiments. The use of BERT represents a novel approach in sentiment analysis for e-commerce by capturing more complex linguistic nuances and improving classification accuracy. The results provide a comprehensive evaluation of consumer feedback, enabling e-commerce platforms to assess customer experiences, identify improvement areas, and adapt strategies to better align with user expectations. In summary, this sentiment analysis tool enhances the online shopping experience by delivering precise evaluations of customer reviews, supporting informed purchasing decisions, and fostering a user-friendly e-commerce environment.Accuracy, precision, recall, and F1 score metrics were used to evaluate the model, demonstrating notable gains in performance. The findings reveal the importance of domain-specific sentiment analysis models, supporting more accurate feedback classification. These results enable e-commerce platforms to assess customer experiences, identify improvement areas, and adapt strategies to better align with user expectations. In summary, this sentiment analysis tool enhances the online shopping experience by delivering precise evaluations of customer reviews, supporting informed purchasing decisions, and fostering a user-friendly e-commerce environment.