基于混合蛙跳模糊聚类的电子商务协同过滤推荐

Collaborative Filtering Recommendation in E-commerce Based on Shuffled Frog Leaping Fuzzy Clustering Algorithm

针对传统协同过滤推荐算法生成推荐速度慢、推荐质量不高等缺陷,提出了一种基于混合蛙跳模糊聚类的改进协同过滤推荐算法。算法首先利用模糊C-均值(FCM)聚类方法对用户数据进行预处理,得到用户数据聚类中心,有效地降低了推荐工作量。然后选取相似度最优的若干聚类组成候选用户邻居集合,并利用混合蛙跳算法快速地全局寻优能力得到用户最近邻居集合,提高了推荐精度。最后,通过计算预测评分生成推荐结果。仿真结果表明,相比于传统协同过滤推荐算法,该算法在推荐速度和推荐精度上有明显改善。

As the traditional collaborative filtering recommendation algorithm （CFRA） having many problems, such as slow recommended speed and low recommend quality, an improved CFRA （ICFRA） is proposed based on the shuffled frog leaping fuzzy clustering algorithm. With the fuzzy C- means （FCM） clustering method, the user data is preprocessed, and the user data clustering center is presented, effectively reducing the workload of recom- mendation. Several clusters having the best similarity are selected to form the user neighbor set, and the shuffled frog leaping algorithm is used based on its fast global optimization ability to get the nearest neighbor set, which im- proves the accuracy of recommendation. Finally, recommendation results are generated by calculating the score. The simulation results show that, compared to the traditional collaborative filtering algorithm, this algorithm has better performance in the recommended speed and accuracy.