UW Research Science Team Takes 2nd in WSDM Cup

Congratulations to Carl Bergstrom, Jevin West, and Ian Wesley-Smith (henceforth referred to as Team Eigenfactor) for taking 2nd Place — and being the top team from North America — out of a competition pool of about 90 research teams in the 2016 WSDM Cup Challenge!

The Cup Challenge is to “assess the query-independent importance of scholarly articles, using data from the Microsoft Academic Graph — a large heterogeneous graph comprised of publications, authors, venues, organizations, and the fields of study. The goal of this ranking challenge is to provide the best static rank values for each publication entity in a heterogeneous graph. Static rank plays a key role in recommendation systems, especially in the cold start scenarios, and also for search engines to determine the ranking of search results. Traditional metrics have relied heavily on citations, which favor the more established, seminal papers and treat all citations as equal (and positive) indicators of importance and impact.”

“The cool part of the challenge,” says eScience Institute Data Science Fellow Jevin West, “is that they provided two different solution sets for the ‘black box.’ This was to prevent overfitting. We chose a method that would be robust to permutations of the data. In fact, we were in 15th place for much of the first data set until there was a permutation on the data set.”

The top eight finishers in the challenge will be invited to write a paper and present their method at the WSDM Conference, taking place this February in San Francisco. The top teams will also work with the Bing team at Microsoft to incorporate their rankings into their academic search engine.

“We’re really pleased by this,” Carl Bergstrom wrote in an email announcing how the team performed. “It shows that the algorithms we’ve been developing as part of the Eigenfactor project compare favorably to machine learning approaches being developed around the world. Entering the nearly parameter-free article-level Eigenfactor algorithm into a supervised learning contest was definitely an instance of bringing a knife to a gun fight. I’m glad our knife was sharp enough that we were able to walk out of there alive.”

You can read more about the WSDM Cup Challenge rules here.

And big congratulations again to Team Eigenfactor!