| Title | : | Algorithms and Infrastructure for Scalable Processing of Graph Data |
| Speaker | : | Srikanta Bedathur (IBM Research) |
| Details | : | Mon, 30 Jan, 2017 12:00 PM @ BSB 361 |
| Abstract: | : | We are witnessing an enormous growth in both the size and the variety
of graphs in domains ranging from online social networks to knowledge
representation. Analytics and query processing over these graphs is
challenging because even simple graph algorithms become inordinately
expensive to evaluate. Addressing this not only requires designing
practically scalable algorithms but also building carefully crafted
storage and processing infrastructure.
I will start this talk by outlining our recent work in both these directions -- ranging from indexing solutions for fundamental graph algorithms to graph mining algorithms over knowledge graphs and query processing engines for graphs. Next, I will present our PathSketch algorithm and its extensions as a specific example of scalable algorithms we have developed. This algorithm can efficiently estimate the shortest path between a pair of vertices in the graph, help in the exploration of alternate paths and handle user-defined label constraints. Then, as an example of an end-to-end data management system we have built for semantic graphs, I will describe Quark-X, a scalable RDF system that enables the modeling of higher-order relationships between entities and confidence scores on the extracted relationships, and support ranking queries in SPARQL. Finally, if time permits, I will briefly list the recent research directions I am pursuing in the area of graph analytics. Bio: Srikanta Bedathur is currently a senior researcher in the Knowledge Engineering and Data Platforms group of IBM India Research Labs and an adjunct faculty member at IIIT-Delhi and IIT-Delhi. Before joining IBM in 2014 he worked at IIIT-Delhi as an Assistant Professor where he established and headed the Max-Planck Partner Group on Large-scale Graphs. Earlier, he was a senior researcher at Max-Planck Institute for Informatics, and simultaneously held the positions of adjunct junior faculty member at the Saarland University and at the Cluster of Excellence on Multimodal Computing and Interaction. He received his Ph.D. from the Indian Institute of Science in 2005. His current research interests are broadly in the fusion of ideas from databases, information retrieval and machine learning, and recently with emphasis on large-scale graph management, graph analytics and visualization. |
