Survey on Anomaly Detection in Web Usage Mining

Abstract: In machine learning, Anomaly detection is widely processed for many real-world applications mainly intrusion or fraud detection which also require an actual and efficient framework to classify differed data instances. In this paper, the study of anomaly detection in web usage mining is done. In which, the various methods and functions that are used in various papers in various methods of analysis. This survey attempts to provide broad overview of the research on anomaly detection. Also have done the work in the form of collected existing techniques into different categories based on the method adopted by each procedure. For every paper the used techniques and what that author process are all explained in various behaviors.

Keywords: Anomaly detection, outlier detection.