REU 2011 Project Title
Visual Mining of Communities in Complex Networks: Bringing Humans Into the Loop
Description of the project
Recent advances in information technology have led to the emergence of a new discipline, called network science, where the goal is to understand network behavior in complex topologies. This project will focus on a key problem -- called "Community Discovery" -- in network science, where the goal is to find the "most appropriate" community structure in a given network. Example networks include social networks, communication networks, and citation networks.
Eliassi-Rad's prior works provide algorithms for detecting community structure from networks  and for visual analysis of networks . A challenge here is obtaining and using feedback from users in the community-discovery process. One aspect of this research is to develop new human-computer interfaces for understanding, navigating, and revising community structures. Another aspect is to better formalize the notion of an "appropriate" community by learning from user-provided examples and constraints.
Requirements: A basic course in probability and statistics, a course on algorithms, and strong programming skills in Java.
 Keith Henderson, Tina Eliassi-Rad, Spiros Papadimitriou, Christos Faloutsos: HCDF: A Hybrid Community Discovery Framework. SDM 2010: 754-765.
 Zeqian Shen, Kwan-Liu Ma, Tina Eliassi-Rad: Visual Analysis of Large Heterogeneous Social Networks by Semantic and Structural Abstraction. IEEE Trans. Vis. Comput. Graph. 12(6): 1427-1439 (2006).