Personalized E-learning Resource Recommendation Method Based On Multi-objective Particle Swarm Optimization Algorithm With Neighborhood Learning Strategy

With the development of information technology,e-learning has become an important way of learning.Compared to the traditional way of learning,e-learning is not limited by time and space and can meet students’ learning needs at any time.The e-learning resources is the basis of e-learning,e-learning resources are abundant and rich and can meet the needs of learners for all kinds of learning goals.However,learners need spend a lot of time to retrieve the screening process,this will make the students lose interest in learning and learning problems.At the same time,the learner’s knowledge level,cognitive ability and learning style and other factors not considered in the process of the existing e-learning resources,increases the difficulty of geting right e-learning resources.In order to solve the above problems,this paper proposes the personalized e-learning resource recommendation method based on multi-objective particle swarm optimization algorithm with neighborhood learning strategy(DNVMOPSO_RA),we parametrically represent the features of learners and learning resources on the network,and construct the e-learning resource recommendation problem model as multi-objective problem,and use multi-objective particle swarm optimization algorithm with neighborhood learning strategy(DNVMOPSO)to optimize the multi-objective problem.Our DNVMOPSO_RA method can match the individual characteristics of learners and the characteristics of e-learning resources,so as to recommend personalized e-learning resources for learners and improve their learning efficiency.The neighborhood learning method is used to improve DNVMOPSO algorithm’s convergence accuracy and diversity.The neighborhood concept comes from the multi-objective evolutionary algorithm based on decomposition method and is integrated multi-objective particle swarm optimization algorithm,so that the particle can learn more neighbor particles.Through the test of multi-objective test function,the experimental results show that the DNVMOPSO algorithm is superior to the existing multi-objective intelligent optimization algorithms in convergence and diversity.Through the e-learning Resource Recommendation problem testing,results show that the DNVMOPSO algorithm has better performance than other multi-objective intelligent optimization algorithm,DNVMOPSO algorithm is more compatible for solving the multi-objective personalized e-learning resource recommendation problem.The main innovations of this research include:(1)Demonstrating the conflict among the sub problems in the multi-objective personalized e-learning resource recommendation problem.Through the variety of function value,demonstrating the conflict between the goals of each sub problem,thus proving rationality of constructing personalized network learning resources recommendation problem as a multi-objective problem.(2)Improve the multi-objective particle swarm optimization algorithm and propose neighborhood learning strategy.Particle swarm optimization algorithm(PSO)has better convergence performance in the course of optimization.The multi-objective evolutionary algorithm based on decomposition uses neighborhood concept to find the nearest neighbor for the individual,and obtains the experimental vector through the crossover and mutation between neighbors,and has better diversity.Through multi-objective evolutionary algorithm based on decomposition method’s neighborhood concept introduced into multi-objective particle swarm optimization algorithm,this paper proposes multi-objective particle swarm optimization algorithm based on neighborhood learning strategy(DNVMOPSO).In this algorithm,the particles can not only learn from the history of individual optimal,but also using the neighborhood particle information,thus the algorithm has good convergence and diversity.(3)This paper applies the DNVMOPSO algorithm in the multi-objective personalized e-Learning resource recommendation problem,and put forward the DNVMOPSO_RA method.Many multi-objective intelligent optimization algorithm are introduced into personalized e-learning resource recommendation problem,and compared with DNVMOPSO algorithm.(4)By decomposition method,we find the optimal individual of the Pareto optimal solution set in the multi-objective intelligent optimization algorithm.Different from monocular intelligent optimization algorithm,a global optimal individual can be found,and the optimization result of multi-objective intelligent optimization algorithm is a set of Pareto optimal solutions,and these solutions can not be directly compared with each other.Through decomposition method,the multi-objective optimization problem can be decomposed into one single objective optimization problem with multiple scalar subproblems,the single objective optimization problem function value can be used to compare the size relationship between the solutions of multi-objective problem,thus,in the parper,the decomposition method is used to find the optimal individual of multi-objective intelligent optimization algorithm,the performance of the algorithm is demonstrated by comparing the single objective optimization problem function value of the optimal individual.