Construction of online course evaluation model from the perspective of learning willingness

With the gradual “desensitization” processing of the online learning platform, it is difficult for researchers to construct student portraits through incomplete learning data. This paper uses the public evaluation and message data of the relevant online courses, uses the Latent Dirichlet Allocation (LDA) to extract the subject words, introduces the topic correlation parameters and course characteristic parameters, and finally constructs the LDA online course evaluation model by visualizing the results. Firstly, taking the “innovation and entrepreneurship” course as an example, it studies how to allocate online course resources more reasonably and effectively through new technologies and new carriers, which are suitable for online course developers, managers and researchers to study the characteristics of a certain type of course. Secondly, through the setting of course characteristic parameters, the model can also be applied to the specific course analysis, this paper takes the “Modern Etiquette” of Hunan University and the “Advanced Mathematics (I)” course of Tongji University as examples for visual analysis, and provides a reference for the course team to carry out teaching intervention and teaching decision-making.