An Enhanced Transaction Identification Module on Web Usage Mining

Abstract

Web data mining is a process which transforms unorganized or less-structured data on the World Wide Web into useful information. Web usage mining, one kind of Web data mining, focuses on exploring users' browsing pattern on a single Website from Web access data restored in the Web server. Web usage mining in general consists of three steps: (1) clean/filter the Web access data. (2) convert the Web access data into a set of transactions by deploying transaction identification modules, and (3) explore the relationship of interest in the converted transactions using data mining techniques. The quality of the relationship being discovered from the transaction set largely depends upon the data conversion process as information might be lost during data conversion, and effective transaction identification modules arc thus necessary. In this study, we propose an enhanced transaction identification module. The simulation results show the effectiveness of our proposed method compared with the early studies.