Research Of Online Learning Resources Sequencing Service Based On Particle Swarm Optimization Algorithm

With the deep integration of artificial intelligence technology and educational application research,the online learning system which is based on teacher guidance,learner dominance and domain knowledge model is constantly improved.Online learning,which has become one of the important internet applications,can not only solve the restricted problem of geographical location,learning time,learning resources and other aspects in traditional learning field,but also can provide learners with convenient,real-time and interactive online learning environment.Computable learning situation,understandable learning subject and learning service customization are the three core issues to realize personalized online learning.According to the differences of learners' cognitive ability,the resource contents and sequences that match learners' abilities and needs are found,imported,combined,generated and distributed to the learners.Providing an intelligent,dynamic and personalized online learning resource sequencing service for the learners as well as improving the learning efficiency of online learners has become an important research content in the field of online intelligent learning.However,in the process of online learning,the personalization features are variable and hard to quantify.The mass and complexity of online learning resources lead to a series of problems in the research of online learning resources sequencing service,such as the difficulty of learning resource recommendation,the slow recommendation,the low accuracy and low learning path matching.In this context,this paper starts with the mathematical model construction of the online learning resource sequencing service in different stages,and analyzes the characteristics of online learning resources sequencing service.Through the optimization of particle swarm algorithm in the field of computational intelligence,this paper improves the matching degree of learning resource recommendation and learning path optimization,promote the learning efficiency of online learners,explores the effective mechanism of online learning resources sequencing service,and builds a complete theoretical system of online learning resources sequencing service.The research work and contributions mainly include:1.A fuzzy adaptive binary particle swarm optimization algorithm based on evolutionary state determination(EFBPSO)is proposed.The algorithm uses fuzzy classification method to divide the evolution state of population into S1 and S2 states.In the S1 state stage,we use the S mapping function and the larger inertia weight value to improve the convergence speed while ensuring the stability of the algorithm.In the S2 state stage,we use V mapping function and dynamic inertia weight adjustment strategy to enhance the algorithm's later global exploration ability,so as to avoid falling into the local optimum too early.Simulation results show that EFBPSO algorithm has better convergence speed and accuracy.The EFBPSO algorithm is applied to the field of online learning resource recommendation,and online learning resource recommendation method EFBPSO-RA is proposed.The recommendation performance of this method is analyzed in detail.2.The novel approach of online learning resource recommendation based on multi-dimensional feature differences is proposed.This paper develops the research of online learning resource recommendation service based on multi-dimensional feature differences,and constructs the online learning resource recommendation model POLMRM based on multi-dimensional feature differences of learners and learning resources.Considering the learners' preference,the collaborative filtering technique is used to predict the characteristic preference of learners in the model.The collaborative filtering algorithm combined with the particle swarm optimization algorithm.The chaotic function is used to initialize the population.The inertia weight adjustment method is combined with the inertia weight linear reduction method of the Hamming distance.The dynamic optimization of the inertia weight parameters and the diversity of the group in the particle swarm optimization algorithm EFBPSO is carried out according to the change of the mean value of the sea distance between each particle and the optimal particle of the group.Then,an adaptive binary particle swarm optimization algorithm ABPSOA is designed.This algorithm is combined with the online learning resource recommendation model POLMRM,and proposes the online learning resources recommended method POLMRM-RA is proposed which is based on the multi-dimensional feature difference to optimize the personalized online learning resource recommendation performance.3.The novel approach of online learning path optimization based on multi-dimensional information feature mapping model is proposed.The research of online learning path planning service based on multi-dimensional information feature mapping model is carried out.On the basis of analyzing the factors such as learning resource difficulty and learner's ability,learning expenditure,learning resource and knowledge matching degree,the multi-dimensional information feature mapping model(MIFMM)for online learning path is constructed.The online learning resource ranking rule is introduced,which is repulsion degree,and online learning resource sequence optimization service is realized according to the repellent degree.On the basis of optimizing the speed updating strategy of EFBPSO,we replace the first half population with the last half population of the smaller Hamming distance.The algorithm improves the global exploration ability and jumps out of the local optimum.An improved binary particle swarm optimization algorithm called NBPSO is proposed.The NBPSO algorithm and the multi-dimensional information feature mapping model MIFMM are combined and applied to the online learning path optimization field.An online learning path programming approach,MIFMM-PPA,is proposed.The performance of the optimization method is analyzed in detail from two aspects that are path generation process and path programming approachc omprehensive performance.4.A novel approach of online curriculum resource generation based on neighbor mean variation multi-target particle swarm optimization is proposed.In the view of multi objective optimization,the research of online curriculum resource generation service is carried out.The multi-objective optimization model of online curriculum resource generation is constructed with the guidance function of knowledge map,and carries out the online learning resource recommendation sub-target conflict analysis.A neighbor mean mutation operator is introduced to improve the convergence performance and stability of the multi-objective particle swarm optimization algorithm.Based on the analysis of the working principle of the velocity free multi-objective particle swarm optimization algorithm,a neighbor mean variation multi-objective particle swarm optimization(AMOPSO)algorithm is proposed.The AMOPSO algorithm is combined with the multi-objective optimization model of online curriculum resource generation.AMOPSO-GA,an online curriculum resource generation approach,which is based on the multi-objective particle swarm optimization algorithm,is proposed.