摘要： Web日志挖掘的一个主要任务是获得用户的浏览模式，这对Web站点的改进和为用户提供个性化服务提供了非常有价值的潜在信息。该文在分析用户访问模式的特点后，提出了Close模式的概念，基于此概念提出了一种挖掘用户频繁访问模式的Close算法。该算法利用频繁访问模式的封闭特性，挖掘出既是频繁的又是封闭的访问模式，在一定程度上减少了下一阶段“寻找最大频繁访问模式”的工作量。用实际数据对算法的性能进行了验证和分析。

关键词: Web挖掘, 频繁访问模式, 访问模式的顺序子集, Close模式

Abstract: One primary task of Web log mining is to discover and identify users’ access patterns, which provides very valuable potential information for the improvement of Web sites and the users’ personalized service. The paper proposes a concept of Close patterns after analyzing the characteristic of users’ access patterns. A Close algorithm for discovering users’ frequent access patterns is proposed based on this concept. The Close algorithm discovers frequent and Close patterns, which relieves the next phrase workload of finding maximal frequent access patterns to some extent, by making use of the Close property of frequent access patterns. The algorithm performance is tested and analyzed by actual datum.

Key words: Web mining, Frequent access patterns, Sequential subsets of access patterns, Close patterns