Professor, University of Texas at Arlington
Prof. Sharma Chakravarthy’s research contributions span several related topics in data management, data mining, data analysis, machine learning algorithms, efficiency, and scalability. Research on active capability through event processing was pioneering as such a capability did not exist although was very much needed. This contribution, adopted by the industry, caught the attention of DoD (Department of Defense) for its potential and transitioned into their systems. His research on stream data processing techniques were also pioneering and addressed the problem holistically. He developed a prototype mavStream to showcase the capability, which is being extended for video situation monitoring. This work resulted in an authored book on this subject.
Chakravarthy’s current work on Multilayer Network (MLN) analysis also blazes a new trail. Although MLNs were defined and used earlier, there were no approaches for modeling complex data sets using MLNs and no algorithms were available for analyzing them. We have proposed EER-based approaches to modeling complex data sets to MLNs and a new “decoupling approach” for MLN analysis. His work on "video situation monitoring" uses stream and complex event processing for automating video processing and minimize human-in-the-loop.
Apart from the above, the nominee has done research on Graph mining and scalability using Map/reduce, (multiple) query optimization in databases, semantic query optimization (under Logic and Databases), optimization of data placement in distributed database, social network analysis, and personalized ranking of web search on large data sets.