*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Title: Graph-based Algorithms, Models and Tools for Security, Healthcare and Finance
Acar Tamersoy
School of Computational Science and Engineering College of Computing Georgia Institute of Technology http://www.cc.gatech.edu/~atamerso
Date: Friday, November 6, 2015
Time: 3:30pm EDT
Location: KACB 1212
Committee:
Dr. Duen Horng (Polo) Chau, School of Computational Science and Engineering (Advisor) Dr. Shamkant B. Navathe, School of Computer Science (Co-Advisor) Dr. Munmun De Choudhury, School of Interactive Computing Dr. Rahul C. Basole, School of Interactive Computing Dr. Kevin A. Roundy, Symantec Research Labs
Abstract:
Real-world graphs (a.k.a. networks) have become omnipresent, existing in a variety of forms ranging from social networks to biological networks. Not only that graphs have become ubiquitous, but they have also grown in size over the past few decades, leading to tera-scale graphs. It has been postulated that such graphs have a strong potential to benefit societies at large in today's big data era, by helping solve real-world problems affecting millions of individuals' daily lives, which has been the main interest of our research.
This proposed thesis is concerned with graph-based algorithms, models, and tools to solve large-scale societal problems, with a focus on:
(1) Propagation-based Graph Mining Algorithms: We develop scalable graph mining algorithms to propagate information between the nodes based on the graph structure. We present three examples: AESOP unearths malware lurking on the Internet; EDOCS detects comment spammers on social media platforms; and MAGE finds both exact and approximate matches for user-specified subgraph queries on large attributed graphs.
(2) Graph-induced Behavior Characterization: We derive new insights and knowledge that characterize certain behavior from graphs using statistical and algorithmic techniques. We present three examples: a characterization of attributes of short-term and long-term abstinence from smoking and drinking; an exploratory analysis of how company insiders trade; and a study on identifying successful investors in the startup ecosystem.
(3) Enhanced Domain Understanding via Interactive Visualization: We design interactive visualization tools to help users quickly explore and make sense of large graphs, and thereby enhance their understanding of the underlying domain. We present AsthmaFlow, an interactive tool that helps clinicians explore and understand the processes involved in pediatric asthma emergency department care.
Our ongoing work extends these three areas of research. We will develop ADAGE, a framework that partitions a timeline of streaming time-stamped edges into structurally mature graph snapshots. Second, we will characterize predictors of short-term and long-term relapse to smoking and drinking. Third, we will design and develop VISAGE, an interactive tool that supports the visual construction of subgraph queries for visually querying large attributed graphs.
