The Issues and Challenges in Data Pre-Processing and Pattern Discovery for Web Mining Algorithms

Abstract

Web mining is an exciting discipline in the area of data mining as well as classification or clustering. Detecting the usage patterns of the users is significant in using tremendous data accessible in the World Wide Web. Web Usage Mining involves distinguishing usage patterns and has numerous pragmatic applications. It concentrates around methods that can possibly predict the behaviors of the users interacting with the web. It is a nontrivial procedure of mining implicit, unknown and conceivably valuable patterns from a huge database. Web mining can be commonly characterized as the application of data mining procedures to extract the knowledge from the text data from the web. Web mining can be additionally sorted as web content that incorporates text, image, audio, and video etc. Web usage includes log files HTTP, app server logs, etc. In this paper, we aim to give a general view on web usage mining and its significance for originators and those interested in e-commerce and website personalization. Further, the phases involved in web usage mining and the challenges involved in data pre-processing and pattern discovery are presented. Also, it illustrates the various applications and tools along with commonly used algorithms for web usage mining are discussed. Finally, it clarifies a few current research issues and gaps in web usage mining such as privacy and scalability.