The National Gallery turns to AI to better predict attendance figures

The National Gallery has announced a project to use artificial intelligence technology to help analyse past visitor experiences in the museum in order to better predict future attendance and visitor engagement.

The National Gallery is a free museum in London which attracts more than six million visitors a year. It comprises of over 2,300 paintings in the Western European tradition, from late medieval times to the early 20th century from artists including Leonardo, Rembrandt and Monet.

“Machine learning and artificial intelligence have huge potential value for helping museums build better insights and develop new kinds of financial sustainability,” the National Gallery’s IT Director Chris Michaels, told Computerworld UK.

The project involved using machine learning algorithms to predict exhibition visitor numbers, which will help the Gallery gain deeper insights into the visitor experience, better plan future exhibitions and even introduce a dynamic ticketing model.

“The project will enable us to offer better value to our audiences by offering different prices for different exhibitions, while also analysing the behaviour of visits," Michaels explained.

The machine learning work was done by specialist vendor Dexibit, which does most of its work using an Amazon Web Services (AWS) cloud technology stack.

Read next: How the National Theatre protects its scripts and stars from hackers.

Michaels, who joined the gallery in April, said the gallery is the first London museum to introduce AI driven analytics. “What we are focusing on is being a data driven business and how we can improve our visitor experience,” he said.

The National Gallery has traditionally struggled to predict accurate visitor numbers to the museum, especially for temporary exhibitions.

"One of the biggest challenges in implementing forecasting is the impact of a visitor’s decision to view an exhibit – everything from the exhibition focus to the weather outside," Michaels said. "A lot of research needs to go into understanding the visitors, attendance and engagement."

Michaels’ admits working for the Gallery has its benefits though, with a rich collection of data gathered over the years to work from. For this project the Gallery combined this historic data with data from its commercial systems, wifi data, website visits and various open data sources, such as tourism figures from the Office for National Statistics (ONS).

Read next: How the Bolshoi Theatre built its online archives of historic documents

The National Gallery wants to lead the way for the cultural sector when it comes to analytics, transitioning from reporting the museum’s own history, to using AI to predict the future success of exhibitions.

“We are right at the beginning of the journey so we are taking the opportunity both aggressively and with confidence for redefining what a great cultural institution means in the 21st century,” Michaels said.

Copyright © 2017 IDG Communications, Inc.

Bing’s AI chatbot came to work for me. I had to fire it.
Shop Tech Products at Amazon