Ranking English Learning Apps from Google Play Store based on User opinion

Today, educational technology has become an indispensable tool in modern education at all levels. For foreign language training, with the learning demand accounting for over 70% of the country's population, educational technology has attracted startups to develop new EdTech products and drawn interest from investors in this sector. The rapid increase in educational technology for language training has made it difficult for users to choose products. It has posed challenges for management in establishing development and distribution standards and guidelines for product circulation in Vietnam. While numerous methods exist for evaluating information systems, the practice of ranking educational technology applications in specific segments according to user perspectives remains relatively unexplored. Evaluation methods based on customer perspectives will provide valuable insights for companies to improve their products, attract investors, and help users make informed choices. In this study, a method for evaluating and ranking English learning applications on Google Play Store is proposed based on a set of criteria improved by Mc Delone, including system quality, scalability, and user satisfaction. Sentiment analysis techniques measure user satisfaction from customer reviews. The study used Python to collect 9,966 user reviews from 21 English learning apps currently available in the Vietnamese market. A deep learning model was applied to measure sentiment from customer reviews and develop a suitable ranking formula. The study also employed evaluation methods using SVM and Decision Tree to compare with this model, showing that the Deep Learning method was faster, with an RMSE of 0.33 and an R2 of 0.78, making it the most suitable model for predicting customer sentiment.