ABSTRACT Advanced technologies, such as artificial intelligence (AI), have been used by teachers to enhance English as a Foreign Language (EFL) learning. However, few studies have employed AI in the making of video dramas to facilitate EFL learning. In this study, AI was used to recognize the authentic context of a drama and provide students with meaningful lexical resources, including vocabulary and sentence structures, to inspire them in their writing. Video-to-text recognition (VTR) and AI methods were used to develop an AI-enhanced video drama (AI-EVD) application for EFL learning, and its effectiveness for improving EFL writing and speaking was investigated by recruiting 77 university students in three groups: an experimental group (EG), which used the app, and control groups, which used the app without AI features (CG1) or participated in conventional learning (CG2). Learning achievement in the EG was significantly higher than that in CG1 or CG2. An analysis of the correlation between learning behavior and learning achievement showed that VTR-generated vocabulary, generative pretrained transformer (GPT)-generated sentences, and a pronunciation correction mechanism significantly improved the students’ performance. Further analysis with a stepwise multiple regression model revealed that the GPT-generated sentences feature was the learning behavior with the strongest effect on writing and speaking skills. Moreover, the AI-EVD application was perceived as being easy to use, helpful, and engaging according to the results of student questionnaires and interviews. Therefore, the AI-EVD application can motivate EFL students, enable them to produce meaningful video dramas, and improve their writing and speaking skills.
Read full abstract