基于用户模糊聚类的个性化推荐研究

The Research of Personalized Recommendation Based on User Fuzzy Clustering

推荐系统是根据用户的历史浏览记录或对项目的评分记录,自动为用户推送需要的信息,完成个性化推荐功能,是信息获取领域非常重要的技术。首先对用户进行模糊C均值聚类操作,将用户分为用户簇。将加权的欧氏距离替换传统的欧氏距离计算方法,在目标用户所在的用户簇内进行协同过滤推荐,得到Top-n推荐集,为用户完成项目推荐。实验结果表明,该方法可以提高推荐精度,减少评分误差,提高推荐质量,优化推荐效果。

The recommendation system is based on the history of the user browsing history,or the score of the project to automatically push the user to meet the needs of the user information for the user to complete the personalized recommendation of the function,it has become a very important information acquisition technology.Firstly,the fuzzy C-means clustering operation is carried out,and the users are divided into individual user clusters.The similarity in the same user cluster is high,and the weighted Euclidean distance is introduced to replace the traditional Euclidean distance calculation method.The target user’s user cluster within the collaborative filter recommended,and finally get Top-n recommended set for the user to complete the project recommendation.The experimental results show that this method can improve the recommendation accuracy and reduce the scoring error.Therefore,according to the actual problem of fuzzy clustering in the distance calculation to improve,you can improve the recommended quality,optimize the recommended effect.