Research On Resource Recommendation Algorithm Based On Students' Knowledge And Ability Evaluation In Adaptive Learning System

Adaptive learning systems are gaining popularity in the field of modern education.At present,the function of most learning systems is still to assist students in learning knowledge,and there are no examples of learning systems aimed at developing students' abilities.Under the trend of development of contemporary society,the adaptive learning system aiming at improving ability will represent the future of educational technology development.In this thesis,intelligent control and data analysis are introduced into the engineering education process of higher education to create an ability-oriented adaptive learning system.The system is based on students' learning motivation as the system input,the learning process as the controlled object,and the subject ability is the output of the system and serves as the feedback parameter of the system and the design basis of the calibration link..There are many key problems in ability-oriented adaptive learning system,and this thesis studies the most basic and important two of them.The first is to establish a scientific,objective and real-time ability evaluation system.And the second is to establish a learning resource recommendation engine based on ability and knowledge.In view of these two key problems,this thesis has carried out the following main work:1.This thesis presents a three-level ability space based on Thinking Layer,Subject Layer and Knowledge Layer,and a knowledge space based on knowledge map technology.2.Based on capability space and knowledge space,a new capability measurement and evaluation method is proposed by using multidimensional Item Response Theory and Bayesian Network.3.Under the support of the measurement and evaluation method,and take the content-based recommendation algorithm as the core,a learning resource recommendation algorithm based on the students' knowledge model and ability model is proposed.This thesis applies the ability-oriented adaptive learning system to the teaching process of a core professional course of automation major in a university.The system conducted a course teaching experiment in the spring semester of 2019,conducted a real-time assessment of students' ability during the learning process,and recommended personalized test resources for students based on their knowledge model and ability model.After several tests,students' performance has been improved significantly,the results and efficies of the ability assessment system also have been recognized by most students.The experimental results show that the ability evaluation method and recommendation algorithm proposed in this thesis are reasonable and effective,and the students' ability evaluation analysis is realized accurately,and the personalized learning resources that can promote the students' ability are recommended according to the evaluation results.