Affiliations: [a] School of Financial Management, Hefei University of Economics, Hefei, China
| [b] Zhengqi Holding Co., Ltd., Hefei, China
Correspondence:
[*]
Corresponding author. Huang Jingjing, School of Financial Management, Hefei University of Economics, Hefei 230011, China. E-mail: hjj2803@163.com.
Abstract: In view of the individual differences in learners’ abilities, learning objectives, and learning time, an intelligent recommendation method for offline course resources of tax law based on the chaos particle swarm optimization algorithm is proposed to provide personalized digital courses for each learner. The concept map and knowledge structure theory are comprehended to create the network structure map of understanding points of tax law offline courses and determine the learning objectives of learners; the project response theory is used to analyze the ability of different learners; According to the learners’ learning objectives and ability level, the intelligent recommendation model of offline course resources of tax law is established with the minimum concept difference, minimum ability difference, minimum time difference, and minimum learning concept imbalance as the objective functions; Through the cultural framework, the chaotic particle swarm optimization algorithm based on the cultural framework is obtained by combining the particle swarm optimization algorithm and the chaotic mapping algorithm; The algorithm is used to solve the intelligent recommendation model, and the intelligent recommendation results of offline course resources in tax law are obtained. The experiential outcomes indicate that the process has a smaller inverse generation distance, larger super-volume, and smaller distribution performance index when solving the model; that is, the convergence performance and distribution performance of the model is better; This method can effectively recommend offline course resources of tax law for learners intelligently, and the minimum normalized cumulative loss gain is about 0.75, which is significantly higher than other methods, that is, the effect of intelligent recommendation is better.
