This is an archived article. For the latest news, go to the Advance Homepage
For more archives, go to the Advance Archive/Search Page.
Shin Focuses on How Computers Communicate
February 28, 2000

When he was a boy, Dong-Guk Shin was fascinated with the question of how computers will communicate with people in the future. He dreamed of creating a computer like the ones in Star Trek - a marvelous piece of technology that could carry on intelligent conversation with humans in many languages.

In the fullness of time and maturity he came to understand the enormity of bringing such a creation to reality. The technological challenges, alone, gave new meaning to the word daunting.

But something of that youthful fantasy did not die as Shin grew into the realities of a grown-up career in academia. The seed of work that occupies him to this day - work that earned him a Chancellor's Information Technology Award last year - took root in the fertile imagination that once toyed with the idea of such futuristic conveniences as a holo-deck and a replicator.

Shin, an associate professor of computer science and engineering, devotes much of his time to a task that sounds much simpler than it is. His aim is to find ways to "make many databases that are becoming increasingly available on the Internet easily comprehensible by, and accessible to, researchers, scholars and students." The enormity of that task is difficult for the uninitiated to understand.

To put it in non-engineering terms, Shin employs the metaphor of cottage industry. Many research laboratories, for instance, may be working simultaneously on the same project. But the labs, on college campuses separated by hundreds or thousands of miles, function separately and uniquely until they are linked by technology. Each is like a cottage unto itself.

Enter the Internet, an invention that, in little more than a decade, has simply revolutionized the way scholars share information. Suddenly the labs are no longer disparate and remote cottages. Suddenly they are nodes of one giant cyber-lab.

That, at least, is the potential. The challenge to making that happen is a technological mechanism that can account for a plethora of technological platforms, some of which may not be entirely compatible. That is the conundrum that occupies the fertile imagination of Dong-Guk Shin today.

Query Interfaces
A native of Korea, Shin earned his B.S.E. degree at Han Yang University in Seoul and his M.S.E. and Ph.D. at the University of Michigan. He joined the UConn faculty in 1986.

Almost immediately, he began wrestling with the question that would drive his award-winning research. How will the Information Super Highway remain super? As more and more people - in and out of academia - flock to get on that highway, what will be the future of information management? And how can bigger volumes of information be made conveniently available to people who do not have (and are not likely to acquire) the kind of technological expertise that he has?

The answer lies in what Shin calls "query interfaces." These adapters make it possible for end-users to quickly surmount the "semantic barriers" that often stand between them and the volumes of information produced in, for instance, "cottage" laboratories. Unless comparable data about research being conducted simultaneously in different laboratories can be conveniently accessed, the research process is effectively thwarted.

The Human Genome Project
After more than two years at UConn, Shin was still exploring these issues in a highly abstract way, advancing ideas he'd been working with since graduate school. Increasingly, however, he was hungry for an opportunity to apply his theories to actual data.

In 1988, while visiting a friend in Washington, D.C., he learned about the Human Genome Project, the massive research effort launched that year to map the entire human genetic sequence. He was captivated. Recognizing that the project would be undertaken simultaneously in laboratories all over the world - laboratories linked by the rapidly evolving Internet - he returned to UConn and solicited the support of the late Professor Claire Berg, a biologist who became his mentor.

Since then, thanks to a series of grants, he has been occupied by the information technology challenges of that single project. His most recent grant is a more than $500,000, three-year grant from the National Institutes of Health, for work on "A Graphical Ad Hoc Query Interface for GenBank," on which he is the sole principal investigator.

The implications of his research are far-reaching. Combining the computer science disciplines of database systems, artificial intelligence and user-system design, the project has implications for how electronic information will be shared for the foreseeable future. It is, indeed, about nothing less than how computers will communicate with people in the future.

Jim Smith