Modern technology, including the World Wide Web, sensor networks, and high-throughput genetic sequencing, has completely transformed the scale and concept of data in the sciences. Data collections for a number of systems of interest have grown large and heterogeneous, and a crucial subset of the data is often represented as a collection of graphs together with node and edge attributes. Thus, the analysis and modeling of large, complex, real-world networks has become necessary in the study of phenomena across the diverse set of social, technological, and natural worlds. The aim of this workshop is to bring together researchers with these diverse sets of backgrounds and applications, as the next wave of core methodology in statistics and machine learning will have to provide theoretical and computational tools to analyze graphs in order to support scientific progress in applied domains such as social sciences, biology, medicine, neuroscience, physics, finance, and economics.

While the field remains extremely heterogeneous and diverse, there are emerging signs of convergence, maturation, and increased awareness between the disparate disciplines. One noteworthy example, arising in studies on the spread of information, is that social media researchers are beginning to use problem-specific structure to infer between social influence, homophily, and external forces -- subjects historically of intense interest amongst statisticians and social scientists. A second, more long-term example is the growing statistics literature expounding on topics popularized earlier within the physics community. Highly complex application domains, such as brain networks, are coming into the scope of the field.

Goals: The primary goal of the workshop is to become an inflection point in the maturation of social network and social media analysis, promoting greater technical sophistication and practical relevance. To accomplish this, we aim at bringing together researchers from applied disciplines such as sociology, economics, medicine and biology, together with researchers from more theoretical disciplines such as mathematics and physics, within our community of statisticians and computer scientists.

The technical focus of the workshop is the statistical, methodological and computational issues that arise when modeling and analyzing large collections of data that are largely represented as static and dynamic graphs. As the different communities use diverse ideas and mathematical tools, we seek to foster cross-disciplinary collaborations and intellectual exchange. The communities identified above all have a long-standing interest in modeling networks, and while they approach the problem from different directions, their ultimate goals are very similar.

The NIPS community serves as the perfect middle ground to enable effective communication of both applied and methodological concerns. Earlier workshops that we organized at ICML 2006 (on “Statistical Network Analysis: Models, Issues and New Directions”, in LNCS collection, vol. 4503.), and at NIPS 2008 (on "Analyzing Graphs: Theory and Applications"), were well attended, with standing audience at many talks. Since then, network modeling has grown to become a regular part of most if not all of the major conferences that are related to NIPS, and workshops held at NIPS 2009 and 2010 were large by the conference standards (100+ attendees) with overflowing attendance. This year, we are aware of a number of new compelling results and ongoing work, particularly in social media analysis, which reflect increasing sophistication and relevance of the field.

We would like to bring together a diverse set of researchers once again to assess progress and stimulate further debate, in an effort to support a continued, open, cross-disciplinary dialogue. We believe this effort will ultimately result in novel modeling approaches, and ultimately in the identification of diverse applications and open problems that may serve as guidance for future research directions.

We welcome the following types of papers:

1. Research papers that introduce new models or apply established models to novel domains,

2. Research papers that explore theoretical and computational issues, or

3. Position papers that discuss shortcomings and desiderata of current approaches, or propose new directions for future research.

We encourage authors to emphasize the role of learning and its relevance to the application domains at hand. In addition, we hope to identify current successes in the area, and will therefore consider papers that apply previously proposed models to novel domains and data sets.