In order to solve the problem of information retrieval unable to obtain the personal characteristic resource ,the paper introduces the personalized recommendation system concept and the frequently-used algorithm in recommend systems, focus explained the process in cooperative filter technology. The techniques makes use of the user_item matrix figure out the similarity between users, combined with the nearest neighbors' information forecast the personal characteristics of the recommended information. Finally, points out the difficult problem in cooperative filter research .%为了解决信息检索中无法获取符合个人需求的特色资源的问题,介绍了个性化推荐系统的概念以及推荐系统常用的算法;重点阐述了协同过滤技术的流程:利用用户项目矩阵计算用户之间的相似性,结合最近邻信息预测出符合个人特性的推荐信息;最后指出了协同过滤研究的难点问题。
展开▼