Graph-based anomaly detection

摘要: Anomaly detection is an area that has received much attention in recent years. It a wide variety of applications, including fraud and network intrusion detection. A good deal research been performed this area, often using strings or attribute-value data as the medium from which anomalies are to be extracted. Little work, however, focused on anomaly graph-based data. In paper, we introduce two techniques for addition, new method calculating regularity graph, with applications We hypothesize these methods will prove useful both finding anomalies, determining likelihood successful within provide experimental results real-world artificially-created