UT Dallas is One of Four Taking Part in eBay University’s AI Competition

eBay already uses machine learning, but wanted to grow its community and future technologists by working with universities. More than 40 students from four schools are participating in the machine learning competition.

The University of Texas at Dallas is one of four schools participating in a machine learning competition held by eBay University. 

More than 40 students from the schools—also including NYU, Stanford, and University at Buffalo—are participating in the competition, either in teams or as individuals. They are applying artificial intelligence to part of eBay’s public listing data in order to make a product catalog.

UTD is noted for its work and research in AI. Recent projects include research using AI to improve control of prosthetics presented at the 2019 IEEE International Symposium on Measurement and Control in Robotic in Houston and a study conducted by UTD’s Center for BrainHealth and Mindcurrent that used AI and Apple Watch activity data to chart emotions.

For eBay, the product catalog competition allows the e-commerce retailer to make sense out of its more than 1.3 billion listings, which are largely unstructured data. It’s also a way to tap into university tech talent from around the U.S.

eBay already applies in-house machine learning to the challenge, but wanted to grow its community and future technologists by working with universities, according to a company blog post. “By working with universities, we hope that it will pique academic curiosity within ML, spur more research in the ecommerce domain powered by a real-world ecommerce dataset, and help us improve our platform,” the post said.

The competition began earlier this month and will run until around March 4, 2020, with the winning team announced March 25. The specific question competing teams are answering is how to identify two or more listings as the same product by placing those listings into a single group, a capability eBay calls Product Level Equivalency.

The competitors will have access to 1 million unlabeled public data listings with around 25,000 clustered by eBay using human judgement. The clustered listing can be further split into three groups—validation set, quiz set, and final submission set—with each set used for different aspects of the competition.

Winning team members will be offered summer 2020 fully paid internships at eBay’s San Jose, California headquarters.

eBay said the competition was a voluntary effort from its workforce across a variety of disciplines, growing from a hallway conversation to a fully launched challenge.

“We sincerely hope that making this real-world dataset available will entice universities and students to explore the ecommerce domain further and come up with novel approaches to solve complex problems that can have a positive impact on customers and sellers alike,” eBay wrote on its blog.

Get on the list.
Dallas Innovates, every day.

Sign up to keep your eye on what’s new and next in Dallas-Fort Worth, every day.

One quick signup, and you’re done.
View previous emails.

R E A D   N E X T