Stanford computer scientists participate in federal study to improve reliability of electronic voting systems

By Mark Shwartz

Stanford University and five other institutions have been awarded a five-year, $7.5 million federal grant to establish a national research center dedicated to designing more trustworthy voting systems in the United States.

The new project, called A Center for Correct, Usable, Reliable, Auditable and Transparent Elections (ACCURATE), is based at Johns Hopkins University and includes researchers from Stanford, the University of California-Berkeley, Rice University, the University of Iowa and SRI International.

"It is not sufficient that election results be accurate," says David Dill, professor of computer science at Stanford. "The public must know that the results are accurate. That can only be achieved by making election processes as transparent as possible."

An authority on computer-based voting equipment, Dill launched the Verified Voting Foundation in 2004; the nonpartisan organization advocates the use of voter-verified paper ballots in electronic voting machines. He and Stanford colleague Dan Boneh are among several co-principal investigators participating in the ACCURATE project. An associate professor of computer science and of electrical engineering, Boneh is an expert in cryptography developing special codes that keep computer data secure.

"One of the issues is how do you know that the right software is running on your computer," Dill says. "Another question is how to maintain the integrity of audit logs that record who accessed data."

In addition to technology experts, the ACCURATE project also includes a psychologist who will study how people interact with computers and a law professor who will examine legal and public policy issues. "Improving our voter system will require a truly multidisciplinary research effort," Dill adds.

The ACCURATE project is part of Cyber Trust, a National Science Foundation program that also seeks to develop new ways to protect the nationwide computer infrastructure.

-30-