Sep62011

Researchers Help Intelligence Community Improve Predictions of Global Crises

Print-Friendly Version

By Catherine Ferraro

What if we had accurately predicted the events on Sept. 11, 2001, or the attack on Pearl Harbor during World War II? It’s a question often considered by the intelligence community as they work to predict and prevent the next global crisis. The accuracy of their predictions has the potential to save lives and even change the course of history.

To help improve the accuracy of these predictions, Mason is leading a multi-institution research team on a project called Decomposition-Based Aggregative (DAGGRE). The team was awarded a $2.2 million contract from the Intelligence Advanced Research Projects Activity (IARPA), the research branch of the intelligence community, to participate in a nationwide, multiyear research project called the Aggregative Contingent Estimation (ACE) Program. The contract has a value of nearly $8.2 million if fully funded over a four-year period.

The DAGGRE project aims to improve the accuracy of intelligence analysts in predicting future events. The researchers are recruiting a diverse group of participants to discover whether certain types of people are better than others at making accurate predictions. Participants will be asked to provide predictions about events and trends in areas such as politics, the military, economics, business and science and technology.

Participants’ forecasts will help the researchers gain more knowledge about new methods that can be used for collecting and combining the opinions of many people to provide more accurate forecasts, as well as to improve the communication of these results.

“The results from the DAGGRE project could change the way we forecast global events and thereby enable decision makers to make more informed choices,” says Charles Twardy, principal investigator on the project and research assistant professor in Mason’s Center for Command, Control, Communications, Computing and Intelligence (C4I).

“The goal is to demonstrate the effectiveness of combining the knowledge of many individuals in a unique way that improves accuracy beyond what any one person or expert could provide,” he adds.

Twardy explains that forecasting methods will range from using bias-reducing questions to advanced artificial intelligence methods that determine individual strengths or notice “bandwagon effects” where everyone reacts to the same piece of news.

“The core of the system is a new method for combining different forecasts from many analysts,” he says.

Hundreds of research questions will be developed and housed on an interactive website. Participants will log in to the website, choose statements about specific events of interest to them, and then estimate how likely these events are to happen. Deadlines for each forecasting statement will be determined by the researchers. Participants may change and update their answers at any time. Eventually, participants will also be able to interact with other forecasters through online social media.

“The research we are doing on the DAGGRE project is extremely important because it can help the intelligence community to reduce its chances of being surprised by events that it should have seen coming,” says Kathryn Laskey, co-principal investigator and professor of systems engineering and operations research and associate director of the C4I Center.

“If our research helps analysts make more accurate predictions, it has the potential to improve our lives in many ways,” she says.

Other universities and organizations that are working with Mason on the DAGGRE project are the Australian Center for Excellence and Risk Analysis at the University of Melbourne, James Madison University, Mercyhurst College, Defence Research and Development Canada at Toronto, nemoSibi Ltd. and KaDSCi LLC.

Mason’s C4I Center is the nation’s first and only civilian university-based entity offering a comprehensive academic and research program in military applications of information technology. The center performs research in areas such as communication and signal processing, command support and intelligent systems, and information systems.

More information about the DAGGRE project can be found on the website.

Leave a Comment

Back to Main Page