Learning Path Recommendation Based On Multilayer Grid-Like Knowledge Model

Under the background of more and more detailed requirements for post competence and higher requirements for professionalization in the employment market,college students who study knowledge content according to the tree knowledge model of "subject-course-chapter-knowledge point" in traditional teaching mode and existing adaptive learning systems are unable to meet the requirements of employers and to be competent for actual production jobs due to the lack of learning and training for professional skills.Aiming at solving the problem of mismatch between college students and actual production jobs in the employment market,this thesis studies the learning path recommendation technology based on multilayer grid-like knowledge model in the adaptive learning system.By constructing reasonable domain knowledge model and learner model,this thesis completes the design and implementation of the adaptive learning path recommendation algorithm to help learners find out the gap between the existing knowledge ability level and the actual production jobs,and complete the knowledge content learning according to the personalized learning paths.The main research work in this thesis is as follows:1)A multilayer grid-like knowledge model is proposed by analyzing of the shortcomings of traditional teaching mode and existing adaptive learning system,and knowledge ontology in computer domain based on multilayer grid-like knowledge model is constructed in this thesis.2)The learner model is designed.Learners' learning style is predicted by analysis of behavior data using Tree Augmented Naive Bayesian Networks(TAN-BN).The mastery of knowledge points is estimated based on Item Response Theory(IRT).Knowledge space model is used to reflect learners knowledge ability level.3)Two criteria for selecting learning objectives are proposed.And adaptive learning path recommendation algorithms are proposed based on these two criteria.These algorithms help learners to make up for the gap between existing knowledge ability level and actual production jobs by planning a reasonable learning sequence of knowledge points for learners.4)The effectiveness and rationality of algorithms are analyzed and evaluated by experiments.The results of experiment show that adaptive learning path recommendation algorithms proposed in this thesis achieves good results.