混合智能算法在电子商务个性化推荐中的研究  

A Study of Hybrid Intelligence Algorithm of Personalized Recommendation in E-commerce

作　　者：刘利民[1] 刘晓莉[1] 

机构地区：[1]内蒙古工业大学信息工程学院,呼和浩特010051

出　　处：《内蒙古工业大学学报（自然科学版）》2011年第3期359-363,共5页Journal of Inner Mongolia University of Technology：Natural Science Edition

基　　金：内蒙古自然科学基金项目(2010MS0913)

摘　　要：针对传统协同过滤推荐算法出现的推荐精度下降、邻近用户缺乏多样性等问题,本文将免疫遗传模糊c-均值聚类算法应用于传统协同过滤推荐算法中以求解决上述问题。免疫遗传算法收敛速度快、搜索精度高、稳健性强等特点与模糊c-均值聚类算法柔性模糊划分的特点相结合,可以快速的收敛到全局最优解,有效的克服了模糊聚类中的弱点,从而提高了个性化推荐系统的推荐质量。According to some issues like the recommended accuracy down and lack of diversity due to applying to the traditional Collaborative Filtering recommendation algorithm,this article discusses how Immune Genetic Fuzzy c-means Clustering Algorithm is applied to the traditional Collaborative Filtering recommendation technology.It is a combination of advantages of Immune Genetic Algorithm and Fuzzy c-means Cluster Algorithm,integrating fast rate of convergence,high searching precision,strong robustness of Immune Genetic Algorithm and flexible fuzzy classification of Fuzzy c-means Cluster Algorithm.It converges optimal solution of the entity and overcomes disadvantages of fuzzy partition,improving the accuracy of recommended results.

关 键 词：电子商务个性化推荐 协同过滤 免疫遗传模糊c-均值聚类 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]