针对目前启发式算法用于解决个性化网络学习资源推荐问题时存在推荐速度较慢、不稳定等问题, 文章提出基于改进二进制粒子群算法的个性化网络学习资源推荐方法~(AsyBPSO-RA). 该方法将个性化网络学习资源推荐问题建构为适应度函数, 利用改进二进制粒子群算法~(AsyBPSO) 优化此适应度函数, 生成推荐结果; AsyBPSO 采用非对称映射函数, 取代基本二进制粒子群算法中的~S 型映射函数, 以更好地平衡算法的探索和开发阶段. 通过五组实验结果对比分析发现, AsyBPSO 收敛能力强, 稳定性高, 表明~AsyBPSO-RA 是较为有效的个性化网络学习资源推荐方法.

In this paper, we propose a personalized e-learning resource recommendation method based on an improved binary particle swarm optimization algorithm (AsyBPSO-RA) to solve the problem of low recommendation speed and instability in personalized e-learning resources recommendation methods with current heuristic algorithms. AsyBPSO-RA constructs the personalized e-learning resource recommendation problem as a fitness function, then uses an improved binary particle swarm optimization algorithm (AsyBPSO) to optimize the fitness function for generating recommendation results. AsyBPSO uses an asymmetric transfer function to replace the S shaped transfer function in the basic binary particle swarm optimization algorithm, for balancing the exploration and exploit phases of algorithm. Through the comparisons of five groups of experimental results, it is found that the convergence of AsyBPSO is strong and the stability is high, which indicates that AsyBPSO-RA is an effective way to recommend e-learning resources.