U.S. Department of Energy

Pacific Northwest National Laboratory

Dr. George Chin

Principal Investigator
Dr. George Chin
(509) 375-2663

Project -- Dynamic Network Analysis via Motifs (DYNAMO)

Dr. George Chin, Ph.D., is a Chief Scientist in the Visual Analytics Group at PNNL. George has brought in and led a wide array of research and development projects in the areas of graph analytics, visual analytics, and machine learning for DOE and DoD clients. His most recent research work is focused on streaming dynamic graph analytics, where he is developing scalable methods for identifying cyberattack patterns in real-time computer network traffic, constructing probabilistic graphical models for tracking potential network attack vectors based on software vulnerabilities, and utilizing innovative temporal and geospatial graph metrics to identify leaders (or rabbits) and followers (or wolves) in co-location networks.

More generally, George has developed and applied machine learning techniques to analyze large-scale dynamic data. In this area, he has developed and applied parallel Bayesian network learning and inference algorithms on various projects, primarily focusing on large-scale temporal and spatio-temporal models.  These algorithms have been applied to support fault detection, pattern recognition, and forecasting and analysis for various types of real-time networks. George has also researched, developed, and parallelized various dynamic graph algorithms for analyzing large-scale, streaming data.  On various projects, he has developed new efficient algorithms to detect precursor events and patterns as they emerge in complex large-scale dynamic networks. This work has been applied to intelligence applications, cybersecurity, distributed sensor networks, and simulations.

George has also studied social, organizational, and work structures, processes, and characteristics as they relate to scientific or analytical research. In one study, George researched and characterized the problem-solving and collaborative processes of scientists from various domains. In another study, he investigated the social, collaborative networks that scientists form while interacting and working in an on-line virtual instrumentation laboratory. Finally, George examined and elaborated various qualities of intelligence analysts’ work such as what investigative methodologies they apply, how they triage information, how they identify patterns and trends, etc. Related research has focused towards constructing analyst-centered representations of analyses through graphical models, scenarios, ontologies, and knowledge bases. The goal is to capture the semantic knowledge and meaning embedded in an analysis in a way that is comprehensible and intuitive to analysts and that may be operationalized to support analytical goals and functions.

George received a B.S. degree in Computer Science from Oregon State University and a M.S. and Ph.D. in Computer Science from Virginia Tech, where he studied under Dr. Mary Beth Rosson and Dr. John Carroll in the area of human-computer interaction.

| Pacific Northwest National Laboratory