Anomaly detection is one of the most challenging and critical issues in information trust and cyber-security research, especially in the face of rapidly growing big data and increasingly skilled and vicious cyber-attackers. In this talk, we will present some of our recent initial studies on this issue and discuss how to perform data mining in heterogeneous information networks for anomaly detection. Our research shows that special attention should be paid to three methodologies for mining outliers in heterogeneous networks: (i) query-based mining of individual outliers or networked outliers, (ii) mining of outliers in static heterogeneous networks, and (iii) mining of outliers in dynamic heterogeneous networks. Although we present some interesting preliminary studies, we believe that the complexity of the problem and the challenges of diverse applications call for in-depth study. Thus, we will leave some time in this seminar for brainstorming and discussion about this interesting frontier.
Jiawei Han is the Abel Bliss Professor of Computer Science at the University of Illinois at Urbana-Champaign. His research focuses on data mining, information network analysis, database systems, and data warehousing. Professor Han has over 600 journal and conference publications on those topics. He has chaired or served on many program committees for international conferences, including service as PC co-chair for the KDD, SDM, and ICDM conferences and as the Geographical Area Coordinator for the Americas for VLDB conferences. He was the founding Editor-In-Chief of ACM Transactions on Knowledge Discovery from Data. He currently serves as the Director of the Information Network Academic Research Center supported by the U.S. Army Research Lab, and as the Director of KnowEnG, which is an NIH Center of Excellence in Big Data Computing and part of the Big Data to Knowledge (BD2K) initiative. He is a Fellow of the ACM and the IEEE, and the recipient of the 2004 ACM SIGKDD Innovations Award, a 2004 IEEE Computer Society Technical Achievement Award, the 2009 IEEE Computer Society Wallace McDowell Award, and the 2011 Tau Beta Pi Daniel C. Drucker Eminent Faculty Award at UIUC. His book Data Mining: Concepts and Techniques has been used popularly as a textbook worldwide.