Dr. Amarnath gupta of San Diego Supercomputer Center of UC San Diego

A polystore is a data platform that sits on top of multiple data management systems that support different data models and provide a query language and a query processing engine that can query across these systems. A polystore is “analytical’’ if, in addition to data stores, it also supports one or more analytical libraries such that a user can freely interleave data retrieval and analysis operations. We have developed an end-to-end analytical polystore called AWESOME that supports a workload specification language called ADIL and operates over relational, graph and text data models. This talk will present the design considerations, architecture, engineering tradeoffs and query optimization in the AWESOME query processor. We will also present a diverse set of applications and workloads, together with some performance results. Finally, we will touch upon some newer results on cross-model query rewriting and optimization issues we are designing in the AWESOME system.

Biography:

Dr. Amarnath Gupta received his PhD in Computer Science from Jadavpur University in India. He is currently a Full Research Scientist at the San Diego Supercomputer Center of UC San Diego and directs the Advanced Query Processing Lab. His primary areas of research include semantic information integration, large-scale graph databases, ontology management, event data management and query processing techniques. Before joining UC San Diego, he was the Chief Scientist at Virage, Inc., a startup company in multimedia information systems. Dr Gupta has authored over 100 papers and a book on Event Modeling, holds 13 patents and is a recipient of the 2011 ACM Distinguished Scientist award.

Email:

Address:Full Research Scientist , San Diego Supercomputer Center of UC San Diego, San Diego, California, United States