Selecting clickstream data mining plans using a case-based reasoning application

Despite the increasing interest and wide use of data mining tools, extracting useful knowledge from the growing available data remains a complex task. The Web environment and clickstream data demand even more efficacy to knowledge discovery: such data is a rich, complex and huge source of information; the discoveries are easily actionable; and time constrains are typically hard. One general, crucial and very challenge issue of knowledge discovery is the selection of the right methods to apply according to the nature of the problem under analysis. This issue, particularly within the Web usage mining scope, is the focus of our work. This paper describes a case-based reasoning system especially oriented to assist users in the development and application of Web usage mining processes. This system takes as inputs the characteristics of the available data, the analysis’ requirements, and, based on acquired experience from successfully processes, delivers a solution: a mining plan suited to the data analysis problem at hand.