Analysis of Customers Interest for Web Log Clustering

The mining of the web is the process of directly obtaining data from the Internet using data mining methodologies and techniques through hyperlinks, server logs, publications on the Internet, web-based services, and other sources. Web mining is the process of gathering and analysing data in order to look for patterns in web content that might provide insights about consumer behaviour, market trends, and human behaviour in general. It is the process of extracting information from the Web that people require. In this case, the information offered by the Web not only matches the precise information that the user requires, but also suggests material that is related to that specific information. Three major tasks of web content, web structure, and web usage mining that are included in the category of web mining. One part of web mining is called “web usage mining,” which gathers information from web log files about the online activity of users and clients. To identify people who behave similarly on a specific website, it must be used to examine web log files. Web usage techniques are analyzed, including the preprocessing steps involved in web mining. This research study presents a click-stream data based on web usage mining approach to determine customer's interest for products, particularly in relation to things sold on e-commerce websites. Then, this article examines the K-Means clustering, DBSCAN and Hybrid Density Based K-Means Clustering after introducing the idea and methodology of data mining as well as the categorization and methodology of Web data mining for web log file. The system show the comparison of successful clustering results for mean squared errors in K-Means and Hybrid Density Based K-Means clustering algorithm.