A Data Mining On Emotion Analysis And Knowledge Difficulty In Chinese MOOC Forum

MOOC provides people with the opportunity to study the world's top universities' courses for free. With the popularity of MOOC, it has generated a large amount of diversified data, this area gradually attracted more and more researchers' interest, but most of them focus on modeling students to predict their learning outcomes or whether the completion of the curriculum, while the MOOC discussion area has not been much researched, especially for the Chinese MOOC platform discussion area.The students in MOOC have various identities and different learning background, and there are problems of low completion rate, so the teaching program needs to be properly modified, usually based on students' work and examination results. Discussion area is an important tool for students and teachers to communicate, to master the mood of the discussion area is also an important part of the assessment of student learning quality. The purpose of this paper is to analyze the mood of the discussion area with the data mining method to select the specific category of discussion posts reducing the teachers' time of browsing and answering the message and extract the knowledge points which is difficult to help the teachers to arrange the course knowledge structure more reasonably and improve teaching methods.In this paper, we use the theory and method of machine learning and natural language processing to construct a data mining system for MOOC forum. In this paper, we divide the discussion posts into three categories:"confused", " explanations" and "unrelated", and then the SVM model is used to predict the discussion posts' categories. The TF-IDF and TextRank algorithms are used to extract the key words for the posts of "confused"and "explanations", which is the preliminary result of the difficult knowledge points.This paper firstly introduces the theoretical knowledge used in the data mining system achieving a good theoretical basis for the system implementation. This part mainly introduces the classification algorithms in machine learning and the keyword extraction algorithms in natural language processing. Then the design and implementation of data mining system are introduced in detail. Finally, this system is used to deal with six courses, we discuss and analyze the results obtained, and compare the results of different courses' similarities and differences.