科研成果详情

In this paper, we propose an English Language Concept Learning App by combining English concept learning and AI technologies such as automatic options generation and speech recognition analysis. The application is based on WeChat mini-program aimed at assisting and conducting English learning by providing various exercises such as multiple-choice and phonetic questions. To generate wrong options for multiple-choice questions, we used new words from WordNet, a large English lexical database with collocations detected by using statistical natural language processing and neural network models such as CBOW in Word2vec or BERT. User voice input is embedded in voice interaction for recognition and analysis. Based on the framework, this application is expected to be combined with other Artificial Intelligence (AI) technologies to analyze users’ performance and adjust subsequent teaching curriculum accordingly.