Communications of the ACM
How Your Location Data Is Being Used to Predict the Events You Will Want to Attend
Researchers at the University of Cambridge are using location-based data from Foursquare to determine the factors that make people more likely to attend one event than another.
Credit: Technology Review
University of Cambridge researchers led by Petko Georgiev are tapping data from location-based social media network Foursquare to learn what makes people more likely to attend one event rather than another.
The team says when the most critical factors are accounted for, they can accurately predict event attendance.
The effort involved the collection of Foursquare data on the movements of about 190,000 people in London, New York, and Chicago over an eight-month period in 2010 and 2011, and the researchers also leveraged data on networks of friends, home locations, and the date and times of the movements supplied by the social network. They mined the data initially to find events, and also estimated the events' social attraction by tallying the number of friends who show up at the same places. To determine how these variables correlate with whether an individual turns up for an event, the researchers used 90 percent of the data as a training set and then employed the remaining 10 percent as a test data set. The presence of friends at an event turned out to be the major factor that influenced attendance, while other vital indicators included distance from home, the time people are active, and previous behavioral patterns.
From Technology Review
View Full Article
Abstracts Copyright © 2014 Information Inc., Bethesda, Maryland, USA
No entries found