Enhancing Directed Content Sharing on the Web

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.

Firehose

Drip

Water Cooler

see everything

carefully limit intake

collaborative filtering

see everythingsee everything

managed time and qualitymiss interesting posts

personalizedmakes errorsrequires training

FriendsourcingRelated to your research

Our goal is to expand the directed sharing process by making it easier and less spammy.

http://feedme.csail.mit.edu

1. Offer sharing for little time and effort2. Surface activity via awareness

indicators3. Learns 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 survey (N=40 / N=100) on Amazon Mechanical TurkVetted for cheaters: none rejected.Paid $0.20 / $0.05

IntroUnderstandingSupportingEvaluationConclusionFe

edM

e

E-mail is still dominant

E-m

ail

Talk

ing

in p

erso

n

Socia

l net

wor

k sit

es

Inst

ant M

essa

ge

Twitt

er

Blogg

ing

plat

form

s

News ag

greg

ator

s

Socia

l boo

kmar

king

Stum

bleU

pon

RSS/Fee

d Rea

der

0

5

10

15

20

25

30

35

40

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

What explains interest in sharing?Sharing: likeliness of sharing web content with friends, family, and colleagues

“I often tell people I know about my favorite web sites to follow. “

Seeking: time and interest spent finding interesting web content

“I often seek out entertaining posts, jokes, comics and videos using the Internet. “

Bridging social capital: weak ties“I come in contact with new people all the time.”

Bonding social capital: strong ties“There is someone I can turn to for advice about making very important decisions.”

[Ellison et al. 2007]

β p-valuefactor

Seeking .74 < .001

Bridging Social Capital

.22 < .05

Bonding Social Capital

.01 .33

Adj. R2 = 0.56

12

34

56

7Sharing

1 2 3 4 5 6 7Seeking

IntroUnderstandingSupportingEvaluationConclusionFe

edM

e

Can we give active content seekers the means to share more?

FeedMe’s target usersFirehose: active information seekers• Purposely consume volumes of

content• Use aggregators like Google Reader

Thimble: recipients• Won’t use a new tool, but read e-mail

RecommendationsAnnotate each post with friends who might be interested in the content

Recommendations

[email protected] [email protected] FeedMe today 0 FeedMes today

[email protected] FeedMes today

Type a name…

Add an optional comment… Now

Later

Lifehacker: Share with friends using MIT’s FeedMe

Awareness indicators

[email protected] 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?”

[email protected] FeedMes today

[email protected] it already

Digests: managing volumeShare without overwhelming the inbox

Now Later

One-click thanksLow-effort recipient feedback

[email protected]

Building models without recipient involvement

[email protected] FeedMes today

[email protected] FeedMe today

MIT HCIResearch

Computer Science

Education

Recommendation details

design: 184tweet: 170web: 79

twitter: 48social: 43friendfeed:

32blog: 25

developer: 23

sports: 200baseball: 150

sox: 132lacrosse: 89workout: 41muscle: 30hiking: 23vitamin: 22

New post:

Friend A:

Friend B:

Does FeedMe help?

60 Google Reader users (46 male) recruited through blogsUsed Google Reader daily for two weeks with FeedMe installedPaid $30

IntroUnderstandingSupportingEvaluationConclusionFe

edM

e

Questions

• Do shared posts benefit recipients? • Are the recommendations useful?• Do the social features address

spam and volume concerns?

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

postLikert scale 1–7, mean 5.1 (σ=1.6)

Study designW

ithin

-subje

cts

Between-subjects

Are the recommendations worthwhile?

Speed, Keyboard-Free

Visual Clutter

Do overload indicators [email protected]

5 FeedMes [email protected]

Saw it already

We asked: What killer feature would get you to use FeedMe more?

We measured: unprompted responses regarding social features

14 of 28 without social features asked for them 3 of 30 with social features asked for them

Social Feature: One-click Thanks

“I would like a way to check how many times someone has liked what I have sent to them, compared to how many items I have shared with them.”

30.9% of shares received at least 1 thanks

Discussion

What have we learned?

E-mail as a delivery mechanism

“I'm pretty conservative about invading people's email space…I worry that they will take ‘real’ email from me less seriously”

E-mail as a delivery mechanism

“Email is a more direct way to communicate, and I feel that articles that are I read are more like 'ambient' information.”

Low-priority Queue

IRCIMMailing ListGoogle ReaderTwitterFriendfeedFacebook

Mixed-initiative Social Recommenders

• Sharers appreciate recommendations• High error tolerance• Low marginal cost to sharers• Applications to other AI-hard

problems– Social search– Expert finding“We think that your friend Sanjay can answer this question about Nikon cameras: […] Is he a good person to ask?”

FeedMe Not Installed: 93.8%

FeedMe Installed: 6.2%

Post Recipients

Bootstrapped Learning

30.9% One-click Thanks

Bootstrap from intersection of recommendations

Privacy

Summary of Contributions

IntroUnderstandingSupportingEvaluationConclusionFe

edM

e

• Formative understanding of the process behind link sharing

• Leveraging social link sharing to power a content recommender

• Users acting as AI gatekeepers for others

MIT HUMAN-COMPUTER INTERACTION

http://feedme.csail.mit.edu

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 < .001

Bridging Social Capital

.22 < .05

Bonding Social Capital

.01 .33

Adj. R2 = 0.56

Hypotheses

1. Sharers seek out large amounts of web content

2. Sharers are especially social individuals

Hypotheses

1. Sharers seek out large amounts of web content

2. Sharers are especially social individuals