Developing a new ensemble method for sentiment analysis in mobile assisted language learning: a case study for Duolingo

In today's world, mobile devices and mobile technologies have become one of the indispensable elements, especially for young people. Learning activities using these technologies have also become widespread, and mobile assisted language learning (MALL) has become even more important. This study was conducted to evaluate users' opinions about MALL methods. For this purpose, Duolingo user comments, which is currently the most known and used mobile application in foreign language education, were used. One million comments to the app are classified in terms of sentiment analysis. In the study, a new model was proposed by combining different feature extraction and classification methods and the results were compared. It has been determined that the proposed model has high classification success. With the proposed model, it is thought that user opinions can be analysed and software and applications can be developed according to user needs, especially for foreign language learning.