NICOP - Massive network analytics
Abstract
Massive network analytics:Recent advances in information and communication technologies have shown dramatic improvement and unparalleled development of our networked world. Almost all humans are interconnected and beyond that, all objects of our environment are becoming networked, forming huge global systems. The objectives of this proposal are to provide scientific foundation for developing algorithms that will allow massive graphs problems to be tackled not only on expensive supercomputing infrastructure, but also on more available, off-the shelf equipment. This will be achieved by addressing four aspects: (1) Sub-linear graph algorithms, (2) Scalable graph algorithms, (3) Analytics for processes on massive networks, and (4) Local versus global metrics for massive graphs. The targeted research goals of the project are to provide scientific foundation for developing algorithms that will allow massive graphs problems to be tackled not only on expensive supercomputing infrastructure, but also on more available, off-the shelf equipment. This can be achieved by considering algorithms that are randomized and return an approximate solution rather than the exact one. The project long term vision is to develop a scientific foundation of data science for massive networks, which in turn will generate a technological leap from 1010 to 1018 in graphs size for networks analysis.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Nov 23, 2016
- Source ID
- N629091612222
Entities
People
- Ljupco Kocarev
Organizations
- Office of Naval Research
- United States Navy