【摘要】 在实际应用中,一个网站的网页对用户有着不同的重要性。该文每个网页由不同的专家给出语义上的重要性评估,这些语义评估再被转化成表示权重的模糊集。此外,由于树结构在挖掘频繁项目时不需要产生频繁项集及对这些频繁项进行测试而被广泛应用于数据挖掘中。然而,单纯的树结构可能造成有用信息的丢失。因此,该文提出了一种新的结构——频繁链表加存取树(简称为FLAAT),并在此结构上提出了一种有效的加权偏爱模式的挖掘算法。试验证明该文的算法是行之有效的。更多还原

【Abstract】 Web pages in a web site may have different importance for users in real application.In this paper,web pages are firstly evaluated by managers as linguistic terms to reflect their importance,which is then transformed as fuzzy sets of weights.Besides,tree structure is extensively adopted in data mining because it doesn’t need to generate the frequent items and test them.However,only tree structure may result in the loss of useful information.Therefore,a novel structure,frequent link and access tree(simplified as FLAAT) is proposed,on which an algorithm is provided to mine weighted preferred patterns from web logs.The experiment shows it is effective.更多还原