The role of sentiment analysis in personalized learning has received increasing attention, but many existing platforms have not yet been able to effectively combine sentiment analysis and adaptive learning. To this end, this study explores the application of computer adaptive technology in an AI-based business English intelligent learning platform, aiming to optimize the learning process through advanced algorithms and improve learning efficiency and personalized experience. The study collects learners’ behavioral data in real time and combines artificial intelligence algorithms to dynamically analyze and model learners’ needs, thereby achieving personalized learning path recommendations and real-time feedback. This study first uses support vector machines and cluster analysis to perform personalized modeling of learners and identify their learning progress and weak links. Secondly, this paper predicts learners’ behavior through collaborative filtering algorithm and automatically recommends suitable learning content experiments. The data results show that the business English intelligent learning platform based on computer adaptive technology significantly improves learners’ learning efficiency and participation. Through personalized recommendations and dynamic adjustment of learning content, learners can master more business English vocabulary, grammar and oral expression skills in a short period of time.