Towards an enhanced evaluation framework for English reading competence: leveraging multimodal learning analytics

As the field of educational assessment is growing, traditional ways of testing English reading ability cannot adequately show all the different kinds of information that learners use when they read. Because of this, how to employ multimodal learning behaviour data to make more accurate assessments is a popular topic in educational research right now. This research suggests the MLB-ERAM model for assessing English reading proficiency based on facts on how people learn in different ways. MLB-ERAM uses a lot of multimodal learning behaviour data and deep learning (DL) technology to get a whole picture of how well students can read. The experimental results reveal that the MLB-ERAM model works well with multimodal data, gets around the problems with standard assessment methods, and is a useful guide for the future growth of educational assessment technology.