Mason team offers insights at National Gallery of Art’s datathon

Mason art history graduate student Paul Albert presents the Mason team's findings at the National Gallery of Art's datathon in October. Photo courtesy of the National Gallery of Art, Washington, D.C.

A George Mason University team was among the teams of art historians and data scientists participating in the National Gallery of Art’s first-ever datathon, “Coding Our Collection.” The NGA is the first American art museum to invite groups to analyze its permanent collection data.

Mason art history graduate student Paul Albert and School of Business marketing faculty members Laurie Meamber and Gautham Vadakkepatt presented their findings on Oct. 25 before an audience at the National Gallery of Art in Washington, D.C. Other teams came from Carnegie Mellon University, Duke University, and the University of California, Los Angeles, among others.

“We were interested in using data science to understand the contours of our collections and how it has evolved over time,” said Sarah Greenough, NGA’s senior curator. “This datathon opens a new door at the gallery—and hopefully other museums—giving us new ways to understand our collections.”

The datathon coincides with other major efforts by the gallery to make its collection more widely available to the public. Using the gallery’s own data, teams looked at diversity in the materials on display at the gallery, diversity in the gallery’s acquisitions over time, and other topics.

The Mason team looked at the popularity of the artists in NGA’s collections. Using Wikipedia page views as a popularity measure, Albert said they assigned each artist featured in the collections an engagement score.

From left, Mason art history graduate student Paul Albert and School of Business marketing faculty members Gautham Vadakkepatt and Laurie Meamber participated in the National Gallery of Art's datathon in Washington, D.C. Photo provided.

While the team wasn’t surprised to learn NGA’s most popular artist was Leonardo da Vinci, “what surprised us was the second most popular artist was Neil Armstrong, the astronaut, whose moon photographs are among the NGA’s collection,” said Paul Albert, who gave the presentation.

Albert said that, of the more than 10,000 artists presented at the NGA, the top 20% of artists accounted for 97% of the total Wikipedia page views. These are artists whose names most people would recognize—Pablo Picasso, Michelangelo, Andy Warhol, and Ansel Adams.

“The engagement scores the Mason team developed can help NGA’s outreach efforts to attract and educate the public,” said Albert. “[The scores] can help NGA leverage the public’s interest and shape its social media efforts.”

What intrigued Albert about the project was the opportunity to examine the idea of “value” when it comes to art.

“The field of art creates an idea of what is valuable,” said Albert. “But what does the general public value?”

Albert, who led the Mason team, came to Mason after retiring from a career in the data sciences. At Mason he divides his time between classes in art history and computational social sciences. He consulted with the Roy Rosenzweig Center for History and New Media on this project and provides Tableau software training on campus.

“Digital humanities offer a new and different way to understand things,” said Albert. “It broadens our understanding of our world and teaches critically important analytical skills.”

He adds: “We are drowning in data. It is vitally important we develop data literacy among all Mason students, including those studying the humanities. Not only to improve their prospects in the job market, but to also improve their ability to be citizens of this new world.”

To view a video of the presentation and interact with the dashboards the team built for the challenge, see