Research on the Topic Mining of Learners' Interest Based on the Mongolian MOOC Platform Course Discussion Text

At present, one of the key directions of MOOC research is to meet the individual learning needs of learners, while the focus of personalized learning is to model learners’ interest in learning, and whether the model can accurately reflect learners’ interest and admiration plays a central role in the lesson recommendation mechanism. This research takes "Introduction to Computer" course of the Mongolian MOOC platform as the research object, and discovers the topics of interest of the learners by digging the content of the course discussion area. First, after crawling the content of the discussion area, the text needs to be preprocessed, including encoding conversion, text proofreading, removing stop words, and removing affixes; secondly, the discussion text is described by the vector space model, and the keywords and their weights are calculated by the TF-IDF algorithm; finally, the semantic similarity of keywords is calculated through the cosine formula, and after clustering, the topics of interest of the learners are obtained. The experimental results show that the learner's reason for choosing a course is related to three themes, namely the content of the course, the teaching method and the learning experience.