Abstract: In the personalized recommendation systems, the resulting association patterns do not perform well in predicting future browsing patterns. This paper proposes a new personalized recommendation method, which is applied to Web log mining by integrating user clustering and association-mining techniques to improve the Web personalized recommendation. The experiment result shows that the method can improve the precision rate and F rate of evaluation criteria effectively.

Key words: Web usage mining, Cluster analysis, Association rule, Personalized recommendation

摘要： 在基于Web使用挖掘的推荐系统中，仅采用关联规则挖掘技术的Web推荐系统在预测用户未来浏览模式时很难取得令人满意的结果。该文将聚类分析方法结合关联规则推荐算法，应用于Web日志文件的挖掘，以改进个性化的推荐方法。实验表明，该算法能够显著地改进推荐测度的精确率指标和综合评价指标。

关键词: Web使用挖掘, 聚类分析, 关联规则, 个性化推荐

CLC Number: 

• TP393.092