BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:The Effect of Recommendations on Network Structure - Sharad Goel (
Stanford University)
DTSTART:20161215T140000Z
DTEND:20161215T144500Z
UID:TALK69513@talks.cam.ac.uk
CONTACT:INI IT
DESCRIPTION: Co-authors: Jessica Su (Stanford University)\
, Aneesh Sharma (Twitter)
\;
Online social net
works regularly offer users personalized\, algorithmic suggestions of who
m to connect to. Here we examine the aggregate effects of such recommenda
tions on network structure\, focusing on whether these recommendations in
crease the popularity of niche users or\, conversely\, those who are alre
ady popular. We investigate this issue by empirically and theoretically a
nalyzing abrupt changes in Twitter'\;s network structure around the mi
d-2010 introduction of its "Who to Follow" feature. We find that users ac
ross the popularity spectrum benefitted from the recommendations\; howeve
r\, the most popular users profited substantially more than average. We t
race this "rich get richer" phenomenon to three intertwined factors. Firs
t\, as is typical of network recommenders\, the system relies on a "frie
nd-of-friend"-style algorithm\, which we show generally results in users
being recommended proportional to their degree. Second\, we find that the
baseline growth rate of users is sublinear in degree. This mismatch betw
een the recommender and the natural network dynamics thus alters the stru
ctural evolution of the network. Finally\, we find that people are much m
ore likely to respond positively to recommendations for popular users --
perhaps because of their greater name recognition -- further amplifying t
he cumulative advantage of well-known individuals.
Rela
ted Links
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR