Discovering roles and types from hierarchical information networks

Graphs are all around us. They can be made to model countless real-world phenomena ranging from the social to the scientific including information, media, biology, chemistry, medical systems, and e-commerce systems. These graphs are called information networks because they represent bits of information and their relationships. In my thesis, I investigate the principles and methodologies for discovering roles and types in large information networks by exploring hierarchies inherent within the networks. I focus on the Web-information network, as well as specialized sub-networks like Wikipedia, where I aim to determine the type of a Web page or Wiki page as well as its position in the type-hierarchy (e.g., professor, student, and course exist within a department within a college) and their relationships to each other. By identifying multiple types and roles such networks become {\em heterogeneous} information networks. This new information can then be used to answer expressive queries on the network and allows for the exploration of additional properties about the network that were previously unknown