The topic of information diffusion, having both theoretical and practical significance, has received attention across several scientific disciplines for some time. However, many empirical studies have lacked the size and accuracy needed to differentiate between theoretical models. This project addressed this problem by serving as a testbed for modeling the dynamics of information diffusion. Beyond testing existing models, this project incorporated network community boundaries and overlapped into descriptive and predictive models of information diffusion and opinion formation.
Online activities leave digital traces, ranging from citation patterns among millions of blogs to avatar interactions in online virtual worlds such as Second Life. This project assembled three very large-scale, high resolution network data sources: one encompassing the interactions of millions of users of the online virtual world Second Life, the second comprising millions of blogs, and the third mapping exact traces with a novel information and opinion sharing application.
These empirical studies of digital traces investigated the effect of social network structure and community boundaries on information diffusion and change. This project also addressed the dynamics of opinion formation in networks of overlapping communities. The research directly fed into network and data analysis courses taught by Professor Adamic as part of social computing and information retrieval specializations in SI. It also added to a collection of online data sets, analysis tools, and tutorials that Professor Adamic shared for educational and research purposes and used to help shape her following studies:
- The Role of Social Networks in Information Diffusion, which used a large-scale field experiment to examine the role of social networks in online information diffusion and found that individuals that were exposed to signals about friends’ information were significantly more likely to spread information and do so sooner than those that weren’t exposed.
- Group Membership and Diffusion in Virtual Worlds, a study that examined social ties in virtual goods transfers and found that individuals a more likely to adopt a virtual good when they belong to the same groups as previous adopters.
- Coevolution of Network Structure and Content, which demonstrated that network structure can be highly revealing of the diversity and novelty of the information being communicated in online settings.
- Limiting the spread of highly resistant hospital-acquired microorganisms via critical care transfers: a simulation study, which examined the extent to which transfers of critically ill patients could lead to the spread of highly resistant organisms. The study also compared the efficiency of different approaches to targeting infection control resources.
- Memes Online: Extracted, Subtracted, Injected, and Recollected, a study that examined the evolution of information, particularly quoted text, as it was collectively processed in social media.
- The Party is Over Here: Structure and Content in the 2010 Election, an examination of Twitter usage by House, Senate and gubernatorial candidates during the 2010 U.S. elections. Findings showed significant differences in the usage patterns of social media between parties and suggested that conservative candidates used this medium more effectively.
- Social Influence and the Diffusion of User-Created Content, a project that used the social network Second Life to identify and model social influence based on the change in adoption rate following the actions of an individual’s friends. This study found that the social network played a significant role in the adoption of content and discovered that sharing among friends occurred more rapidly than sharing among strangers.