Research On Educational Resource Recommendation Service Based On Learner's Personality

Nowadays,with the rapid development of educational informatization,E-learning and M-learning have become important means of learning.Learning activities can be carried out anytime and anywhere,and learners can access various forms of online educational resources in various ways.However,with the explosive growth of online educational resources,the problems faced by learners such as "information overload" and "learning lost" are becoming more and more serious.Therefore,the recommendation service of educational resources has gradually become a research hotspot in recent years.Educational resource recommendation aims at provide intelligent resource push service for learners,which makes learning activities change from the traditional single mode of "people looking for resources" to the intelligent bidirectional mode of "people looking for resources and resources looking for people".It can effectively alleviate the problems of "information overload" and "learning lost" caused by massive resources.It has attracted more and more researchers' attention and become an important research topic in the fields of educational informatization and artificial intelligence.Learners' differences in learning objectives,knowledge levels,learning paths,and learning styles can lead to different needs for educational resources,which makes educational resource recommendations extremely complicated.To some extent,the system must guide learners to complete their personalized learning process.Therefore,how to accurately acquire learners' knowledge level,recommend learning paths suitable for learners' personality and design corresponding recommendation methods of educational resources according to learners' personality have become urgent problems in the field of recommendation of educational resources.This thesis takes the 3rd-grade mathematics education in primary school as the research entry point,based on the basic theories in personalized education,studies the learners' personality mining method,and designs the corresponding educational resource recommendation method.It is of great theoretical value and practical value for improving the accuracy of educational resource recommendation and improving learning efficiency.The main research contents of the thesis are as follows.(1)The learner's personality model and educational resources model are constructed.The basic concepts of learner's personality are analyzed and elaborated in detail.The method of learner's personality mining is studied.The dynamic adaptive learner's personality model and the education resource model based on fuzzy logic are constructed.(2)A knowledge level acquisition method based on cognitive diagnosis model is proposed.In view of the lack of deep mining of the test results,the existing learning tests are unable to diagnose the learners' internal cognitive structure and acquire the learners' knowledge level accurately.Using the rule space model of cognitive diagnosis theory,the process of compiling diagnostic tests for the third grade mathematics learning in primary school is carried out,and the validity of the tests is evaluated.The results show that the rule space model can effectively diagnose the cognitive structure of the learners for the third grade mathematics subtraction operation in primary school.(3)A learning path recommendation method based on Bayesian network is proposed.In view of the shortcomings of the existing path recommendation algorithm,adaptive learning paths are generated by using Bayesian network,which provides a basis for the evaluation of educational resources based on learner's personality.Experimental analysis was carried out in the real learning environment to verify the effectiveness of the proposed method.(4)A learning resource recommendation method based on learner's personality is proposed.Learning resource recommendation needs to comprehensively consider the learner's learning objectives,knowledge level,cognitive ability,learning style to meet the learner's personalized learning needs.At the same time,the classification of learning resources is accompanied by a certain degree of fuzziness.The relationship between learning resources and knowledge points is not whether there is a relationship,but the depth of the relationship.The fuzzy logic method is used to model learning resources.In addition,the learner's knowledge level also has a certain degree of ambiguity.The degree of learners' mastery of knowledge points can not be simply defined as mastery or not mastery,but to what extent.Therefore,the fuzzy cognitive diagnosis model is used to update the learner's knowledge level,and according to the learner's learning objectives,learning path and learning style,the learning resources with high relevance are recommended to the learner.The experimental results on real datasets show that the proposed method is superior to the other three classic methods.At the same time,the effectiveness of the proposed method is further verified by the analysis of practical application results.(5)A new exercise recommendation method based on learner's personality is proposed.In view of the problems existing in the existing methods of exercise recommendation,such as whether the recommended exercises meet the learner's learning objectives,whether the difficulty of the exercises is consistent with the learner's knowledge level.An exercise recommendation method based on learner's personality is proposed.This method takes into account the problem of knowledge coverage and hierarchical structure of knowledge points,updates learner's knowledge level by using job feedback model,and then recommends appropriate difficult exercises according to learner's knowledge level.The experimental results on real datasets show that the method is superior to the other two classical recommendation methods in Precision,Recall and F1.Finally,the effectiveness of the exercise recommendation method is verified by the analysis of practical application effect.In view of some key problems in the recommendation service of educational resources,this paper conducts systematic research,constructs the corresponding learner personality model and educational resources model,put forward the method of learner personality mining,designs the learning resources and exercise resources recommendation algorithm based on learner's personality.The innovative work of this thesis is summarized as follows.(1)A learner's personality model and an educational resource model are constructed.Aiming at the problem of learner's personality dynamic update,a dynamic adaptive learner personality model is constructed.The corresponding educational resource model is constructed by using fuzzy logic method to solve the problem of fuzzy division of educational resources.(2)Mining methods of learner's personality are put forward.The basic connotation of learner's personality is elaborated in detail,and personality mining methods such as knowledge level acquisition based on cognitive diagnosis theory and learning path recommendation based on Bayesian network are proposed.(3)Recommendation methods of learning resource and exercise resource based on learner's personality are designed.It can make up for some deficiencies in the recommendation of existing learning resources and exercise resources.It provides a reference method for the recommendation of educational resources of other disciplines in the field of basic education.The research results can be extended to the design of recommendation systems in other related disciplines or fields.