OSU’s Sharda, Wilson’s research article
recognized as third-most cited paper

by Terry Tush 
(May 3, 2016 at 12:00 pm)
Ramesh Sharda and Rick Wilson

Ramesh Sharda and Rick Wilson

When Rick Wilson arrived at Oklahoma State University in 1990, he was fortunate to begin working with veteran faculty member Ramesh Sharda. The pair teamed up on a research project that ultimately led to the journal article, “Bankruptcy Prediction Using Neural Networks,” which appeared in the top journal Decision Support Systems in 1994.

Little did the now longtime OSU Spears School of Business professors know that more than 20 years later the journal article would be recognized as one of the most cited neural network in business research papers ever published.

The management science and information systems professors’ work was recently listed as the third-most cited paper in the past two decades, according to findings in an article titled, “Artificial Neural Networks in Business: Two Decades of Research,” published by Michal Tkac and Robert Verner in Applied Soft Computing (Volume 38, January 2016, pages 788-804).

“I’ve spent most of the last 20 years here at OSU in administrative positions, so I haven’t focused on research projects as I did ‘back in the day’,” said Wilson, who has served as the Department Chair for the MSIS department since its founding in 2002. “It was a shock to see our paper so high – yet we still see new papers citing it as the basis for future on-going research papers. It is rewarding to know that Ramesh and I played a substantive role in this line of important practical research in the analytics field.”

When the article was first published in 1994, it immediately garnered attention as it was the first research paper that used a sound and thorough experimental design to study the usefulness of neural networks from the artificial intelligence field. “Neural networks are simply brain-inspired statistical tools, but few understood their potential 20-plus years ago,” Wilson said.

The “Bankruptcy Prediction Using Neural Networks” paper systematically analyzed neural network statistical capabilities and showed that it could discriminate features of bankrupt and non-bankrupt firms better than the existing traditional tools. It was a case of artificial intelligence being taken out of the lab and into practice.

In 1990, Sharda and OSU doctoral student Marcus Odom presented a paper on exploring the use of neural networks in bankruptcy prediction problems at the International Join Conference on Neural Networks. That paper is cited in the third-most cited paper in major journals published four years later by Sharda and Wilson.

“It is fun to be exploring the use of new technologies, and gratifying to receive such recognition,” Sharda said. “Of course, having students like Marcus and colleagues like Rick makes it more possible to do high quality work.”

Today, the use of neural networks as an analytics or data science tool is relatively common, but the OSU professors’ paper in 1994 was one of the first to show its potential use.

“Sometimes the general public may not understand business school research,” Wilson said. “Our neural network work is an example of how research on new tools and processes in the academic setting can have a direct and positive impact on business practice. And clearly, this type of research finds its way into the classroom, better preparing our students. This is the land grant mission exemplified. ”

Wilson and Sharda are just two of many outstanding researchers in the Spears School’s data science area of study.

Dursun Delen, the William S. Spears Chair in Business Administration and Patterson Foundation Chair, has had a major impact in this area, as well as in health analytics. He and Sharda have collaborated on a number of books in the data science area.

Rathin Sarathy, Ardmore Chair and Management Information Systems graduate coordinator, has a patent that details data masking algorithms. Bryan Hammer, assistant professor of MSIS, also targets research in the analytics area.

The read the full article and view the rankings of the most cited papers, visit this website: http://www.sciencedirect.com/science/article/pii/S1568494615006122.