College of Liberal Arts and Sciences at Iowa State University

Language barrier

Gary Leaven's research goes hand-in-hand with an honor from the IEEE Computer Society
When Gary Leavens, professor of computer science, looks at computer programming issues, he's
generally not looking to see what’s right with the code.

Instead Leavens is more concerned with what’s wrong in the computer language.

"The long term goal of my research is to understand better how to solve programming problems," he said.

Leavens looks at how to specify such problems, methods for thinking about such problems, notations for expressing solutions, ways to check that the solutions are correct, and tools for automating and assisting with these problems.

"The most useful aspects (in this area) are when we can tell people something is wrong with their program," Leavens said.

Leavens' research in this area is funded by two separate National Science Foundation grants totaling over $300,000. He is collaborating with individuals at MIT, Virginia Tech, and Clemson and in Holland on these projects, as well as Don Pigozzi, professor of mathematics at Iowa State.

He works in the intersection of two areas - programming languages and formal methods - to accomplish his research goals.

Leavens has worked on the design and semantics of object-oriented languages with generic functions.

Such languages are known as multi-method programming languages. One result of this is an extension of Java, called MultiJava, which has some of these features.

His work in formal methods has been focused on ways to specify and verify object-oriented software components. Formal methods apply methods to a program to see what it does and then to find out the mistakes in the program.

The specification work involves the design and formal description of behavioral interface specification languages (BISLs), which record information about detail design.

"The potential impact of the work in formal methods is possibly great," Leavens said. "It might lead to the engineering of software, instead of hacking.

"But more realistically, I view my research as trying to formally understand what one needs to think about when documenting and reasoning about a program or program component," he continued. "This can be of great value for teaching and for the construction of tools, even if people do not use the formalism directly or on a daily basis."

Leavens says these types of programming tools aren't commonly used.

"I'm not sure why that is," he said. "My assumption is that they are too difficult to work with. We're trying to make them easier to use. Along the way some interesting technical issues have come up."

It's this area of research that Leavens feels attracted the attention of the IEEE Computer Society, the largest professional organization of computer science academics.

Leavens has been selected for a three-year term as an IEEE Computer Science Distinguished Visitor through 2005. That program is a popular offering of quality speakers serving IEEE Computer Society professional and student chapters. Leavens anticipates he will give future presentations to Computer Society chapters throughout college campuses in the Midwest.

"This program is a good way to get speakers for the IEEE student chapters," he said. "I think my research is an excellent topic to discuss at such meetings."