Visually spotting patterns in planetary scale graphs

• Grant number: 14/07879-0
• Support Opportunities: Scholarships abroad - Research Internship - Master's degree
• Start date: October 01, 2014
• End date: March 31, 2015
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: José Fernando Rodrigues Júnior
• Grantee: Hugo Armando Gualdron Colmenares
• Supervisor: Duen Horng Chau
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Institution abroad: Georgia Institute of Technology, United States
• Associated to the scholarship: 13/03906-0 - Visualization and processing of planetary-scale graphs using vertex-centric high performance techniques, BP.MS

Abstract

The so-called planetary-scale (or Web-scale) graphs comprise millions of vertices and billions of edges. Such graphs have appeared in several contemporaneous contexts containing information that can potentially explain social behavior (Twitter and Facebook), commercial activities (Amazon), system structures (like roads and computer networks), non-structured information (Wikipedia), information organization (hypertext), to name a few. These graphs are of great analytical value because, besides having a wide range of information, the current trend is that more of such datasets will appear in the following years. For this reason, it becomes necessary the development and the mastering of techniques capable of working with such volumes of information. In this work, we propose investigating very large graphs by means of visual and analytical techniques to understand the behavior of structural patterns according to specific domains, leading to decision making (evaluation, prediction, and identification of alternatives). (AU)