January 28, 2005
4-5:00, Informatics 107

Information Dynamics in the Networked World

Lada Adamic

Abstract:
The shift of communication to the internet, in particular to email, weblogs, and online communities, presents an opportunity to study the information dynamics of social networks on a large scale. We analyze email traffic to obtain an evolving view of the underlying social network structure, which in turn allows us to track the flow of information within organizations. We find that, with few exceptions, information does not spread as a virus would, but rather reaches a limited target audience. We use these same email networks to test how suitable the structure of social networks within organizations is for the classic small world experiment, where one attempts to reach an individual through a chain of acquaintances. Studying one organization, we find that the balance between connections adhering to the formal organization hierarchy and informal communication between different groups is just right to allow individuals to guide their queries through intermediaries.

Weblogs, also referred to as blogs, are another medium well suited to the study of information flow. Some blogs are online journals of individual authors, while others are topic based and may be authored by many individuals. Because blog entries are time stamped, we can not only observe the popularity of information over time, but also identify and visualize likely paths the information takes through the complex network of interlinking blogs. Where the path by which information spreads is ambiguous, we utilize an inference scheme that takes advantage of data describing historical, repeating patterns of "infection." These implicit links can then be used by an influence ranking algorithm, iRank, to identify blogs that are early sources of popular information.

Biography:
Lada Adamic is a research scientist in the Information Dynamics Lab at HP Labs, studying the effects of local interactions on global properties of complex networks. This has led her to study the Internet and the Web, peer-to-peer systems, social networks, and bioinformatics. She received a PhD in Applied Physics from Stanford University in 2001, with her thesis focusing on dynamic processes such as growth and search on the Internet. She has a BS degrees in Physics and Engineering and Applied Science from Caltech.