A Kennesaw State University computer science professor’s invention will allow more robust analysis of text and web search data, enabling researchers to utilize vast amounts of previously hard-to-quantify information into more user-friendly data.

Ying Xie, professor and assistant department chair of the College of Computing and Software Engineering, and Vijay V. Raghavan, the Alfred and Helen Lamson Endowed Professor in Computer Science at the Center for Advanced Computer Studies at the University of Louisiana at Lafayette, have been issued a U.S. patent for a new algorithm on text and web search technology for a “system, method and computer program product for information sorting and retrieval using a language-modeling kernel function.”

“Experimental studies showed that this new search method outperformed traditional search technique,” Xie said. “This method also provides a solid foundation for implementing personalized text or web searches, so it has great value for commercialization.”

Xie, who also teaches in the Ph.D. program in analytics and data science, is a co-inventor for another U.S. patent entitled “Semantic Relationship Extraction, Text Categorization and Hypothesis Generation” that was issued in 2009. He has two more pending patent applications on mobile and social network techniques.

“This latest patent protects an invented technology on text and web search,” Xie said. “Professor Raghavan and I have designed a theoretically sound and practically effective model for information retrieval that is able to integrate evidences of relevance from different sources into a unified framework towards the user’s query.”

Instead of analyzing structured data from neat columns and rows of numbers arranged on spreadsheets, this new search algorithm allows researchers to utilize unstructured data, such as that which exists in email, tweets or texts, for example. They can also perform analysis on video and audio files.

“Professor Ying Xie’s work is on the cutting edge of computing,” said E.K. Park, dean of the College of Computing and Software Engineering. “It pushes the boundaries of deep data even further.”

The ramifications are huge, according to Jennifer Lewis Priestley, director of the Ph.D. program in analytics and data science.

“This means his algorithm, which is basically a series of equations to accomplish a single objective, optimizes the data-searching process. It holds tremendous promise in the area of text mining,” Priestley said. “To put it in layman’s terms, suppose a legal team is preparing for trial. A roomful of lawyers may take weeks to pore through a seemingly endless trove of documents highlighting pages with all the phrases and type them into an electronic database to build their case. Ying’s invention replaces those lawyers because the documents can be scanned by his proprietary algorithm, which would automatically analyze them by key phrases.”

Xie, who is based on Kennesaw State’s Marietta Campus, began teaching in the Ph.D. program in analytics and data science this fall. The program, one of the first of its kind in the nation, received approval earlier this year by the Board of Regents of the University System of Georgia.