ABSTRACT
The analysis of web user behavior has long been a trending topic in the research field of data mining. This is due to the steady increase in internet users and likewise the enormous amount of data that are found and collected in the or server log in the form of clickstream as users browse the internet. In order to analyze web user behavior, web usage mining is used. But a significant amount of time is consumed during the analysis of the web data; hence the user’s behavior can be retrieved by analyzing the clickstream of the users that are classified as highly interested in the website. This is because online user click behavior is practically motivated by their interest. This chapter seeks to classify and cluster users into three different level of interest; highly interested, averagely interested, and less interested following information gotten from NASA dataset. Before the classification, the data is preprocessed to extract useful features and clean all the irrelevant records that might affect the results and also to obtain a structured data within which users and sessions can be easily identified.
