Premium
Intelligent personalised exercise recommendation: A weighted knowledge graph‐based approach
Author(s) -
Lv Pin,
Wang Xiaoxin,
Xu Jia,
Wang Junbin
Publication year - 2021
Publication title -
computer applications in engineering education
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.478
H-Index - 29
eISSN - 1099-0542
pISSN - 1061-3773
DOI - 10.1002/cae.22395
Subject(s) - novelty , computer science , graph , boosting (machine learning) , function (biology) , point (geometry) , recommender system , artificial intelligence , information retrieval , mathematics , theoretical computer science , psychology , social psychology , geometry , evolutionary biology , biology
As a critical function for intelligent tutoring system services, personalised exercise recommendation plays an important role in boosting the study performance of students. However, recent studies on personalised exercise recommendations have only considered the ability of a student during recommendation and have failed to include the essential relationships between knowledge points, which provide a suitable learning sequence of these knowledge points during a study procedure. In this study, we propose an intelligent exercise recommendation method (weighted knowledge graph‐based recommendation [WKG‐R]) for students, based on weighted knowledge graphs, wherein each node represents a knowledge point weighted by the ability of a student and an arrowed edge between two knowledge points indicates their prerequisite relationship. The novelty of WKG‐R can be summarised as follows: (1) It makes a leading attempt to quantify the ability of a student based on various testing behaviours and (2) it attempts to employ the ability of a student and the prerequisite dependencies between knowledge points for enhancing the effectiveness of personalised exercise recommendation. A real classroom teaching practice was conducted to evaluate the effectiveness of the proposed WKG‐R method. The experimental results demonstrated the distinct advantage of WKG‐R in improving the testing scores of students as compared with contemporary solutions. The average score improvement ratio of students for WKG‐R is up to 33%, whereas for the state‐of‐the‐art solution, it is only 22%. The questionnaires collected from the students also reflected a higher level of satisfaction towards WKG‐R than with the contemporary solutions.