Sentiment Identification System for E-Commerce Mobile App Reviews Using Single Layer Neural Network

In the technological era, e-commerce offers business opportunities, particularly through the simplicity of the process of buying or selling products through the Internet. The upkeep of the customer experience must be recognized by e-commerce service providers as a top priority for businesses. Customers can access the global market, compare prices across regions, and even easily compare the services of various e-commerce apps. Online customer reviews on e-commerce mobile apps play an important role, which can be used as personal recommendations for other customers. Because customers rely on the opinions of other customers, negative reviews from customers will deter potential users from downloading the e-commerce mobile app in the future. The system described in this paper uses a single-layer neural network to automatically predict and analyze customer sentiments from online customer reviews. The proposed sentiment identification system model achieved the best performance among the algorithms; it attained an overall sensitivity of 96.2%, specificity of 93.8%, accuracy of 95.0%, and MCC of 0.90. Additionally, the researchers developed a fast and reliable web-based system for identifying sentiment from customer reviews.