Recognizing learners' personality traits based on programming behaviors and its application explorations

Abstract:

How to perform individualized programming teaching is an important issue in educational practice, and how to recognize the personality of learners is the key for this issue. In this paper, the method of recognizing personality is proposed based on the programming behaviors of learners. Specifically, the programming behavior features of learners are first extracted from multiple aspects; then the classification models are established using the support vector machine; finally, the multi-task voting strategy is used to comprehensively identify learners' personality traits. The results show that the traits in the Big Five Model can be predicted using the proposed method, verifying the feasibility of using programming behaviors to identify learners’ personality to a certain extent; in addition, this paper also discusses the applications of the proposed method in future programming education.