A predictive algorithm developed by University of Maryland machine learning researchers to improve Facebook’s notification system for blood donations prompted 5% more people to roll up their sleeves in a test last year , an increase that could mean 140 million more lifesaving donations worldwide if applied to all. of the social media platform’s 2.8 billion users.
Their findings are detailed in a working paper presented at the Association for Computing Machinery’s Economics and Computing Conference.
“This is a successful step to better align global blood supply and demand,” said co-author John Dickerson, an assistant professor of computer science with a post at the Institute for advanced computing from the University of Maryland.
More than 4.5 million Americans need a blood transfusion each year, and the demand is even greater in poorer countries. To help increase supply, Facebook has designed a tool that notifies potential donors of local giving opportunities. Since the tool’s inception in 2017, more than 100 million people in 27 countries have signed up to receive blood donation notifications.
To maximize the tool’s success, Facebook worked with researchers at the University of Maryland Center for Machine Learning to improve its notification process.
Rather than sending random notifications of nearby donation opportunities, the UMD team has developed specialized algorithms to identify whenand or users were most likely to donate, based in part on predictors such as how often they use Facebook, age, or proximity to donation locations.
Then, in the first large-scale algorithmic matching study of its kind, the UMD team tested their predictive model against one that randomly selects nearby donation opportunities. Of the 1.5 million Facebook users who received a notification during the pilot experiment, their prediction exceeded the random model, which is about 75,000 additional donations, the researchers said.
While this potential impact is impressive, there are still many avenues for improvement, said Duncan McElfresh Ph.D. ’21, the paper’s lead author.
These include several models specially designed for different regions or countries, as well as greater coordination with blood collection centers and hospitals, he said.
The team, which includes co-author Karthik Abinav Sankararaman ’16, Ph.D. ’19 on Facebook, is currently preparing an extended version of the paper for a leading academic journal, Dickerson said.