Post on 23-Feb-2016
description
Michael Bernstein, Adam Marcus, David Karger, Rob MillerMIT CSAIL
MIT HUMAN-COMPUTER INTERACTION
Enhancing Directed Content Sharing on the Web
Information Overload
You want more information.
Aggregate
Filter
Facet
Recommend
Friendsourced content sharing
Related to your research
Related to your research
Friendsourced content sharingis inhibited.
Our goal is to encourage friendsourced content sharingby making it easier and less inhibited.
http://feedme.csail.mit.edu
1. Recommend recipients to reduce the time and effort for sharing
2. Surface activity via awareness indicators
3. Learn personalized models passively
• Introduction• Related Work• Understanding Sharing• Supporting Sharing• Implementation• Evaluation• Discussion• Conclusion
Related work• Mediating our information access
– Information mediators [Ehrlich and Cash 94]– Contact brokers [Paepcke 96]– Technological gatekeepers [Allen 77]
• Information is shared via e-mail [Erdelez and Rioux 00] to educate and form rapport [Marshall and Bly 04]
• Recommender systems focus on discovery [Resnick et al 94, Joachims et al 97]
• Expertise recommenders focus on information needs [McDonald 00]
• The FeedMe namesake [Burke 09, Sen 06]
What drives social sharing?
Two surveys (N=40 / N=100) on Amazon Mechanical TurkVetted for cheatersPaid $0.20 / $0.05
IntroUnderstandingSupportingEvaluationDiscussionFe
edM
e
E-mail is still dominant
Talkin
g in p
erson
Socia
l netw
ork sit
es
Instan
t Mess
age
Twitte
r
Blogg
ing pl
atform
s
News a
ggreg
ators
Socia
l boo
kmark
ing
Stumble
Upon
RSS/Fe
ed Rea
der
010203040
Which tools do you use regularly to share web content?
Recipients want more
When asked to agree/disagree with:“I would be interested in receiving more relevant links.”
Median = 6
1 2 3 4 5 6 7
1. Sharers are those who seek out large volumes of web content
2. Sharers are especially social individuals
Hypotheses
What explains interest in sharing?Sharing
“I often tell people I know about my favorite web sites to follow. “
Seeking“I often seek out entertaining posts, jokes, comics and videos using the Internet. “
Bridging social capital“I come in contact with new people all the time.”
Bonding social capital“There is someone I can turn to for advice about making very important decisions.”
[Ellison et al. 2007]
4 scales of 10 questions each
β p-valuefactorSeeking .74 .001
Bridging Social Capital
.22 .05Bonding Social
Capital.01 .33 Adj. R2 = 0.56
<<
1. Sharers seek out large amounts of web content
2. Sharers are especially social individuals
Hypotheses
IntroUnderstandingSupportingEvaluationDiscussionFe
edM
e
Can we give active content seekers the means to share more?
RecommendationsAnnotate each post with friends who might be interested in the content
Recommendations
msbernst@mit.edu rcm@mit.edu1 FeedMe today 0 FeedMes today
karger@mit.edu5 FeedMes today
Type a name…
Add an optional comment… Now
Later
Lifehacker: Share with friends using MIT’s FeedMe
Awareness indicators
rcm@mit.edu0 FeedMes today
Address concerns about volume:“How much are we sending them?”
Give an indication of whether it’s old news“Oh, somebody already sent it to them?”
rcm@mit.edu5 FeedMes today
rcm@mit.eduSeen it already
Digests: managing volumeShare without overwhelming the inbox
Now Later
One-click thanksLow-effort recipient feedback
Implementation
rcm@mit.edu
Building models without recipient involvement
MIT HCIResearch
Computer Science
Education
MIT HCIResearch rcm@mit.edu
Computer Science
Educationrcm@mit.edu
FeedMe Profile
Recommendation details
design: 184tweet: 170web: 79
twitter: 48social: 43friendfeed:
32blog: 25
developer: 23
sports: 200baseball: 150
sox: 132lacrosse: 89workout: 41muscle: 30hiking: 23vitamin: 22
joe@sixpack.com:
rcm@mit.edu:twitter: 38tweet: 30social: 27post: 23
conversation: 19answers: 10blog: 3google: 1
What impact does FeedMe haveon friendsourced sharing?
Two-week study for $3060 Google Reader users (46 male) recruited through blogsUsed Google Reader daily for two weeks with FeedMe installedViewed 84,667 posts; shared 713
IntroUnderstandingSupportingEvaluationDiscussionFe
edM
e
2x2 Study design• Recommendations (within-subjects)
• Awareness and feedback (between-subjects)
vs. vs. vs.
vs.
Do shared posts benefit recipients? • Surveyed 64 recipients, who reported
on 160 shared posts• 80.4% of posts contained novel
content• Appreciative of having received the
post
1 2 3 4 5 6 70
1020304050
Post Ratings
Are the recommendations worthwhile?
Speed, Keyboard-Free
Visual Clutter
Do overload indicators help?rcm@mit.edu
5 FeedMes today rcm@mit.eduSaw it already
We asked: “What killer feature would get you to use FeedMe more?”
We measured: unprompted responses regarding social inhibition
14 of 28 without awareness+feedback features asked for them 3 of 30 with awareness+feedback features asked for them
One-click thanks
30.9% of shares received a thanks
Discussion
Mixed-initiative social recommender systemsE-mail as a delivery mechanism
IntroUnderstandingSupportingEvaluationDiscussionFe
edM
e
Mixed-initiative social recommenders• Humans filter recommendations
for their friends• Small marginal cost:
sharers have already read the article
AI Friend Recipient
Mixed-initiative social recommenders• Sharers appreciate recommendations• High error tolerance
• Applications to other AI-hard problems
[Bernstein et al. UIST ‘09]
Low-priority Queue
E-mail as a delivery mechanism“I'm pretty conservative about invading people's email space.”
“I feel that articles that I read are more like ambient information.”
Summary of contributions• Formative understanding of the
process behind link sharing• Leveraging social link sharing
to power a content recommender• Users as lightweight
recommendation verification for others
http://feedme.csail.mit.edu
http://bit.ly/CHIProgram2010
Study designW
ithin
-sub
ject
s
Between-subjects
38
39
FeedMe Not Installed: 93.8%
FeedMe Installed: 6.2%
Post Recipients
Bootstrapped Learning
30.9% One-click Thanks
Topic relevance drives enjoyment
Questionable content quality
It's awkward
I sent too much already
Too much effort
Might have seen it already
Unsure of relevancy
0 2 4 6 8 10 12 14
What is the biggest concern you have when sharing?
Topic relevance drives enjoyment“Those who know my politics usually send me very pointed articles – no junk.”
“I could care less about a cat boxing.”
Seekingx 10
Sharing x 10
Bridgingx 10Bondingx 10
Verify scale agreementnormality assumptionshomoscedascicityfactor loading
Multiple regression on sharing index
12
34
56
7Sharing
1 2 3 4 5 6 7Seeking
β p-valuefactor
Seeking .74 < .001Bridging Social
Capital.22 < .05
Bonding Social Capital
.01 .33
Adj. R2 = 0.56
Hypotheses1. Sharers seek out large amounts
of web content2. Sharers are especially social
individuals
Hypotheses1. Sharers seek out large amounts
of web content2. Sharers are especially social
individuals
FeedMe’s target usersSharers: firehose• Purposely consume volumes of
content• Use aggregators like Google Reader
Recipients: drip• Won’t use a new tool, but read e-mail
Learn from intersection of recommendations
Privacy