IU researchers participate in LinkedIn project to tackle economic, employment challenges

  • June 24, 2015


BLOOMINGTON, Ind. -- Indiana University researchers are participating in the LinkedIn Economic Graph Challenge, an open call for researchers, academics and data-driven thinkers to propose how they would use data from LinkedIn to solve some of society’s most challenging economic problems.

The IU team's research will focus on the "macro-evolution" of industries, with the goal of forecasting economic trends and guiding professionals toward promising careers.

Yong-Yeol Ahn, assistant professor in the IU Bloomington School of Informatics and Computing, and four doctoral students are among 11 teams to win the chance to participate in the challenge, which seeks to bring the Silicon Valley-based company’s vast amount of business and professional data to bear on solving current economic challenges.

The goal of the IU project, "Forecasting large-scale industrial evolution," is no less than mapping the "entire global economy," according to Ahn. And they’ve got until the end of the year to do it.

"We're really trying to look at the big, big picture," he said. "We’re a bit unique in the breadth of our goal."

But the IU team is up to the challenge. An expert on large-scale network analysis, Ahn has previously developed mathematical models to map complex systems such as the brain, social networks and culture.

The new study will focus on four areas: mapping the connections between industries across the globe; mapping the flow of people between these industries; identifying patterns in people’s mobility; and predicting future employment trends.

"We’re not only trying to give everyone a 'map,' or a large-scale picture of all the sectors of the economy, but also a 'weather forecast' and 'tools,'" Ahn said. "So users will not only know where they can find jobs but also predict bright spots on the horizon -- the emerging markets -- and acquire the skills, or tools, they will need to get where they want to go."

To identify emerging markets, the IU team will draw upon previous research from Ahn and collaborators that found human capital flows toward areas of the greatest opportunity.

"People tend to go to nice places," Ahn said. "We expect workers with the most talent and skills will pool in the hottest new parts of the economy."

To explore an intelligent approach to eliminating skill gaps, the IU team will map skills that LinkedIn users list in their profiles and identify those which seem to transfer most easily to these hot new economic sectors, as well as those that do not offer clear advantages in terms of economic mobility.

If the team discovers a rising number of LinkedIn users list running a cash register among their skills, for example, but also that few users who list this skill move into higher-paying jobs, their algorithm will advise against acquiring the skill, despite the rising demand.

In addition to the potential to improve economic futures for millions of workers, Ahn said the IU team was drawn to LinkedIn by the opportunity to explore truly massive datasets. Currently, LinkedIn reports over 364 million users, including 115 million across the United States.

"This is probably the largest and richest data set that I’ve ever worked on," Ahn said. "The ways you approach data at this level, and tools you use to analyze it, are completely different. We’re using a lot of cutting-edge technologies and thinking in exciting, different ways."

Other members of the IU team are Yizhi Jing, Azadeh Nematzadeh, Jaehyuk Park and Ian Wood, all doctoral students in the School of Informatics and Computing. The project is completely student-driven, according to Ahn, who applied to the contest only after receiving strong interest from the students.

The IU team's mentor at LinkedIn, who will devote his professional time to the IU-led project, is an IU alum. Michael Conover, a senior data scientist at LinkedIn, is a graduate of the IU Bloomington School of Informatics and Computing, where he served as a member of a research group focused on analyzing how information -- and misinformation -- spreads across social networks.

In addition to working with their mentor during the May 11 and 12 orientation session at the LinkedIn headquarters in Mountain View, Calif., the IU team met with members of the other 10 winning teams and absorbed the workplace culture and operating practices of a leading technology company.

"The students were pretty psyched up from visiting the headquarters and seeing how a company of the type of LinkedIn really works," Ahn said. "It's very valuable for our students to gain experience with those sorts of companies."

The IU researchers have until the end of the year to complete the study.

In addition to producing novel research, Ahn said partnering with LinkedIn could put the IU team in position to get their discoveries translated into new products and services, potentially bringing the benefits of IU research to hundreds of millions of job-seekers across the globe.

Other teams selected for the LinkedIn Economic Graph Challenge include members from Carnegie Mellon University, Duke University, the Federal Reserve Bank of New York, Google, Harvard University, the Massachusetts Institute of Technology, Stanford University and the University of Pennsylvania. Their projects focus on topics such as applying machine learning to user profiles, identifying skill gaps, analyzing connections between managers and employees, and reducing inequality in the labor market. 

IU LinkedIn team

From left, Azadeh Nematzadeh, Michael Conover, Jaehyuk Park, Yong-Yeol "Y.Y." Ahn, Yizhi Jing and Ian Wood at the LinkedIn Headquarters. | Photo by LinkedIn

Print-Quality Photo

Media Contacts

Kevin Fryling

  • Office 812-856-2988
  • kfryling@iu.edu