Skip to main content

Feature Stories

Measuring Our Social World
The Computational Social Science Initiative skillfully links science and technology
Computational model on computer screen against a grid of the world

"Our group typifies the kind of broad-scale, interdisciplinary awareness and skill sets needed to fuel the next generation of discoveries and methodological innovations."

- Andrew McCallum

Tracking disease outbreaks. Encouraging energy conservation. Fighting crime and terrorism. Confronting climate change. Strategizing for survival in a globalized economy. Creating suitable educational models for an era of relentless, breakneck technical innovation. Finding ways to make health care both adequate and affordable.Urgent and daunting challenges, every one of them. And they and any number of others now faced by social scientists and policy-makers are all subject to one looming peril: that those who are seeking solutions to them will be unable to master the massive amount of data needed to do the job, but will instead either never find it or choke on it.

To the rescue comes UMass Amherst’s Computational Social Science Initiative. This cross-disciplinary, collaborative group addresses the challenges and opportunities presented by collecting, storing, and analyzing large-scale data related to the social world. Its core faculty members are drawn from computer science, political science, sociology, and statistics. They offer the expertise needed to create practical solutions to problems in quantitative social science—that part of the field devoted to the statistical, mathematical, or computational analysis of data on specific persons or groups of persons in a given country or area.

“Computational social science is particularly important now, as the nation strives to address so many simultaneous challenges,” says Andrew McCallum, the computer science professor who first proposed the initiative and now heads it. “Doing so will require significant improvements in our understanding of the interactions of people, institutions, markets, and the other components that make up our complex social, political, and economic systems. Without substantial innovation in the study of these systems, policy makers and the public will be flying blind as they confront the dramatic changes we’re likely to see in the next several decades.”

Andrew McCallum, Computer Science
McCallum sees the initiative as giving UMass Amherst a leading position in this nascent but rapidly expanding juncture of science and technology—one which, he says, “is poised to revolutionize the social sciences, expand the reach and impact of computer science, and enable our nation to understand the complex systems and social interactions we must manage in order to address fundamental national and international challenges. Immense volumes of data about social and group behavior are now gathered and available, providing the social sciences with information that was inconceivable as little as ten years ago. We’re working to make that data maximally available and comprehensible.”

The Computational Social Science Initiative was inspired in part by Microsoft chairman Bill Gates’s assertion in a 2009 speech that “when you engage in a goal-oriented activity such as development, progress can only be made when you measure the impact of your efforts.” Gates urged that statistics for measuring and modeling development efforts be developed and then form the basis for making policy aimed at bettering the human condition.

This initiative gives the campus national standing as a center of excellence in computational social science and bridges the gaps among the social sciences, computer science, and statistics, inspiring synergy and collaboration. Because of its enormous utility to federal agencies, the military, and private corporations, it offers tremendous fundraising potential.

And, not least of all, it has attracted a host of stellar faculty members. Among them are James A. Kitts, a recently recruited, nationally prominent sociologist interested in the dynamics of cooperation and competition among organizations and among their members; computer scientist David Jensen, who studies the statistical aspects and architecture of systems for knowledge discovery in databases and the assessment of those systems for government and business applications; Hanna Wallach, another computer scientist, who develops mathematical models and computational tools to analyze vast quantities of structured and unstructured data in order to evaluate complex social processes; statistician Krista Gile, who develops methods for sampling data gathered through social networks; sociologist Ryan Acton, who studies the collection and analysis of social behavioral data from web-based sources; and political scientist Bruce Desmarais, who focuses on interdependent political processes, especially political networks and collective decision making.

In Desmarais research, for example, he seeks to understand how interdependent relationships among political actors (e.g., government officials, political organizations, nations) evolve and configure to produce important large scale socio-political outcomes. In one collaborative project, studying more than 12,000 transnational terrorist attacks committed between 1968 and 2002, Desmarais shows how patterns in the evolving network of attacks can be leveraged to forecast which countries will become sources and targets of transnational terrorism in the near future. In another project working with Wallach, Desmarais is using public record archives of intra-governmental email messages to understand and improve the ways in which information flows within government organizations.  

“These people and the others in our group,” says McCallum, “typify the kind of broad-scale, interdisciplinary awareness and skill sets needed to fuel the next generation of discoveries and methodological innovations.”