Research of Data Preparation Based on Web Log Mining
Xia Qi-shou · Journal of Anqing Teachers College · 2008
The basic idea of Web log mining is to apply the technology of data mining to the data source of Web usage.Data preprocessing plays an important role in the field of Web usage mining.The data source of Web log mining is mainly composed of Web logs.This paper introduces detailed content of Web logs.In the light of features of Web logs,some special processing methods of preprocessing are introduced.At the same time,a newly algorithm based on frame-filtering and the access time threshold is proposed at the stage of transaction recognition.Our experiments show that this algorithm can effectively reveal new interesting page groups.