POSTECH professor makes search engines smarter
.jpg?w=728)
Professor Hwang Seung-won
By Bahk Eun-ji
Professor Hwang Seung-won at the Pohang University of Science and Technology (POSTECH) has been working to make Internet search engines more intelligent with the use of big data.
When Internet users try to search for someone or something popular, the results come flying back instantly.
However, queries with vague terms are often automatically reformulated into complicated ones that take longer to provide results.
Hwang and her team have sought to use big data to come up with solutions to respond more effectively and more quickly to various types of users’ queries.
Achieving a consistently fast response time, regardless of the obscurity of the search term, is a challenging goal for her.
To tackle this challenge, she has joined with Microsoft Research.
Microsoft Research has been running a talent training platform over the last 10 years to support joint research between business and academia in Korea.
It has been running more than 200 research projects to support the program at schools for students in the engineering and IT fields.
Hwang’s research is to make Microsoft’s search engine Bing smarter.
“The goal of the collaborative project is to improve Bing search results,” Hwang said in an interview.
She said that even a few queries that take too long to process can undermine user satisfaction. It can also have a negative impact on the revenue of search engine operators.
Hwang graduated from the Korea Advanced Institute of Science and Technology (KAIST) in 1998, and began to teach computer science at POSTECH in 2005.
Hwang said her interest in big data led her to dig into the project.
In their research on how to reduce the latency in returning results, researchers in the team must predict whether a query takes a long time to process and needs extra resources, such as selective parallelization.
Hwang’s team has developed techniques that first identify and then quicken the time needed to answer queries that take more time to reply, thereby improving server throughput by more than 70 percent in experimental trials.
For example, by using past query logs, the team has developed a predictor that spots tail queries with a rate of accuracy of more than 98 percent.
Those time-consuming queries are then handled by a resource manager that the team has perfected, which allocates additional hardware resources to the troublesome queries.
These new techniques have been presented at top-tier conferences, including SIGIR 2014 and WSDM 2015, where the work received a runner-up award.
Hwang said the research project with the Microsoft Research team has allowed them to solve important problems involving search functions for Microsoft and the entire IT industry.
“This technology will eventually provide optimized search results whatever the individual needs, especially in the forthcoming Internet of Things (IoT) era,” Hwang said.