A method for constructing a subject specific curriculum knowledge graph(SSCKG)is proposed to ad-dress issues such as the lack of integration of cognitive theory,incomplete coverage of courses and knowledge points,and unclear display of knowledge correlations in smart learning.Firstly,different data sources are selected for different disciplines and specialties,and on this basis,the data domain and information extraction range are determined by manual screening.The SSCKG pattern layer is formed by using the Stanford Seven-step method combined with Ausubel's cognitive theory and specialty characteristics.Furthermore,according to the above cognitive theory,weight coefficients such as courses,knowledge units and knowledge points are introduced and self-built corpora are completed to expand the curriculum capacity of SSCKG and fill the data layer.Then,according to the above cognitive theory,the concepts of strong correlation and weak correlation between knowledge are introduced and the weight of correlation is set to further distinguish the connections between knowledge and form the knowledge correlation layer to provide support for subsequent personalized recommendation and intelligent examination.Finally,the construction method of SSCKG is formed.In order to verify the effectiveness of this method,the curriculum knowledge graph(SEKG)of software engineering major in Hebei University of Economics and Business was constructed.The practice shows that the method effectively combines Ausubel's cognitive theory and solves the problems such as incomplete coverage of courses and knowledge points and unclear knowledge demonstration.Meanwhile,the knowledge graph can be used to build intelligent question answering system,personalized recommendation system and intelligent examination system to provide support for intelligent learning.
维普
万方数据