Mutual Information to Interpret the Semantics of Anomalies in Link Mining

摘要

This paper aims to show how mutual information can help provide a semantic interpretation ofanomalies in data, characterize the anomalies, and how mutual information can help measurethe information that object item X shares with another object item Y. Whilst most link miningapproaches focus on predicting link type, link based object classification or objectidentification, this research focused on using link mining to detect anomalies and discoveringlinks/objects among anomalies. This paper attempts to demonstrate the contribution of mutualinformation to interpret anomalies using a case study.