Communications of the ACM
Mining Social Networks to Predict Your App Choices
"We thought, 'Can we use social networks to find which apps people might find interesting?'" says MIT's Wei Pan, a Ph.D student at MIT's Media Lab.
Photo courtesy of Wei Pan
Researchers at the Massachusetts Institute of Technology's (MIT's) Media Lab have analyzed the smartphone use of students on social networks in an attempt to determine whether they could forecast which apps they would download.
The project involved 55 postgraduate students, and their smartphone use was monitored over a period of five months. The team created software to identify the most significant friends of each student, note the apps those friends were using, and determine the probability that a given app is owned by a user, based on what their closest friends owned.
"We thought, 'Can we use social networks to find which apps people might find interesting?'" says MIT's Wei Pan. The researchers say that determining the importance of each friend was one of the project's more difficult aspects. During testing, their system correctly predicted 45 percent of the apps on each person's list, compared with about 10 percent accuracy for complete guesses.
The team believes the research could be useful for developers who want to gain insight into why people choose to download certain apps.
From New Scientist
View Full Article
Abstracts Copyright © 2011 Information Inc. 
, Bethesda, Maryland, USA
No entries found