U.S. Department of Energy

Pacific Northwest National Laboratory

Big Graph Search and Analytics: A Journey of Usability and Scalability

Wednesday, April 1, 2015
Dr. Yinghui Wu
Real-life graphs are heterogeneous and huge. These bring two challenges to emerging graph data applications: How to make big graphs usable and useful? And how to adapt the data processing scale to big graphs? In this talk, Dr. Wu will share his experience improving the usability and scalability for big graphs, focusing on graph querying. Topics that will be covered include: • Potential answers are hard to capture with predefined similarity metrics even for simple queries. We suggest transformation-based search to identify top answers. In particular, Dr. Wu will present ontology-based transformation, which harvests external ontologies to interpret query semantics. • How to cope with multiple transformations? Dr. Wu will introduce schema-less searching, which integrates various transformations to a ranking model using machine learning techniques, leading to top ranked matches. • To make graph querying scalable, Dr. Wu will present a distributed querying scheme with provable performance guarantees on network traffic and response time, independent with graph size and partition strategy. These lead to user-friendly graph analytics systems that enable easy data access, search and exploration, and scale well over the growth of data.
Speaker Bio

Dr. Yinghui Wu is an assistant professor at Washington State University. Before he joined WSU, he was a research scientist at Department of Computer Science, University of California Santa Barbara, and a member of the Network Science Collaborative Technique Alliance. His research interests mainly focus on big data, graph databases, graph analytics, and network science, with applications in social/information network analytics and network security. He received his Ph.D. from the University of Edinburgh, UK in 2010.


| Pacific Northwest National Laboratory