CMU-CS-15-126
Computer Science Department
School of Computer Science, Carnegie Mellon University


CMU-CS-15-125

Exploring and Making Sense of Large Graphs

Danai Koutra

August 2015

Ph.D. Thesis

CMU-CS-15-126.pdf


Keywords: Data mining, graph mining and exploration, understanding graphs, graph similarity, graph matching, network alignment, graph summarization, compression, pattern mining, outlier detection, anomaly detection, attribution, culprits, scalability, fast algorithms, models, visualization, social networks, brain graphs, connectomes, VoG, FaBP, DeltaCon, DeltaCon-Attr, TimeCrunch, BiG-Align, Uni-Align

Graphs naturally represent information ranging from links between webpages, to friendships in social networks, to connections between neurons in our brains. These graphs often span billions of nodes and interactions between them. Within this deluge of interconnected data, how can we find the most important structures and summarize them? How can we efficiently visualize them? How can we detect anomalies that indicate critical events, such as an attack on a computer system, disease formation in the human brain, or the fall of a company?

To gain insights into these problems, this thesis focuses on developing scalable, principled discovery algorithms that combine globality with locality to make sense of one or more graphs. In addition to our fast algorithmic methodologies, we also contribute graph-theoretical ideas and models, and real-world applications in two main areas:

  • Single-Graph Exploration: We show how to interpretably summarize a single graph by identifying its important graph structures. We complement summarization with inference, which leverages information about few entities (obtained via summarization or other methods) and the network structure to efficiently and effectively learn information about the unknown entities.
  • Multiple-Graph Exploration: We extend the idea of single-graph summarization to time-evolving graphs, and show how to scalably discover temporal patterns. Apart from summarization, we claim that graph similarity is often the underlying problem in a host of applications where multiple graphs occur (e.g., temporal anomaly detection, discovery of behavioral patterns), and we present principled, scalable algorithms for aligning networks and measuring their similarity.

We leverage techniques from diverse areas, such as matrix algebra, graph theory, optimization, information theory, machine learning, finance, and social science, to solve real-world problems. We have applied our exploration algorithms to massive datasets, including a Web graph of 6.6 billion edges, a Twitter graph of 1.8 billion edges, brain graphs with up to 90 million edges, collaboration, peer-to-peer networks, browser logs, all spanning millions of users and interactions.

230 pages

Thesis Committee:
Christos Faloutsos (Chair)
William Cohen
Roni Rosenfeld
Eric Horvitz (Microsoft Research, Redmond)

Frank Pfenning, Head, Computer Science Department
Andrew W. Moore, Dean, School of Computer Science

Return to: SCS Technical Report Collection
School of Computer Science

This page maintained by reports@cs.cmu.edu