统一事件Web挖掘模型

讨论了电子商务中现有数据收集方式的缺点。基于Web日志的推荐算法具有复杂而不精确的预处理步骤，而客户端收集数据涉及到用户隐私问题。提出了统一事件模型、基于应用层的日志记录，进行用户、会话识别算法和多维关联规则算法UEMFP。指出了UEM在Web挖掘中的优点。实验表明，UEM下的Web挖掘算法具有更好的结果和性能，它可以很好地运用到电子商务应用中。

Unified Events Web Mining Model

Data collection methods used in current E-commerce applications has shown many shortcomings.In Web-log-based methods,many efforts have been spent on the preprocessing of Web logs.Some use client-side logging,which users may reject for privacy reasons.In this paper,an UEM (Unified Events Model) algorithm is presented,which facilitates user and session identification and Web mining models.A multi-dimensional association-rule mining algorithm UEMFP is also discussed,through which we thoroughly discuss the advantages of applying this model to Web mining.The experiments show that algorithms under UEM have better performance and results.It can be well applied in E-commerce applications,including Web personalization,recommendation and business-intelligence.