jeffrey heer · 16 april 2009
DESCRIPTION
Social Computing. Jeffrey Heer · 16 April 2009. Administrivia. Research project abstract drafts, due this Friday, April 17 @ 7am Content Research Question (in one sentence!) Hypothesis Methodology Study Recruitment Plan 1-2 paragraphs (be concise, but concrete ). google earth. flickr. - PowerPoint PPT PresentationTRANSCRIPT
stanford hci group / cs376
http://cs376.stanford.eduJeffrey Heer · 16 April 2009
Social Computing
Administrivia
Research project abstract drafts, due this Friday, April 17 @ 7am
Content Research Question (in one sentence!) Hypothesis Methodology Study Recruitment Plan
1-2 paragraphs (be concise, but concrete)
space
tim
e
asynchronous
co-located
synchronous
co-located
asynchronous
remote
synchronous
remote
projectors
ambient displays
virtual workspaces
webemail
IMtable-top interactio
n
whiteboards
usenet
telephoneteleconference
snail-mailpost-it notes
tagging
flickr
blogs
youtube
graffiti
google earth
distributed visualization
Social Computing [via Wikipedia]
The intersection of social behavior and computational systems.
(a) Supporting social behavior through computational systems
Blogs, e-mail, IM, social networks, wikis Is this different from CSCW?
Social Computing [via Wikipedia]
The intersection of social behavior and computational systems.
(a) Supporting social behavior through computational systems(b) Supporting information production and “computation” by groups of people
Collaborative filtering, prediction markets, tagging, games with a purpose
“The Wisdom of Crowds”
USENET [Smith, Fiore]
World of Warcraft
World of Warcraft [Yee, Ducheneaut et al]
History Flow
Wikipedia History Flow [Viégas et al]
Group Lens
GroupLens / MovieLens [Univ. Minnesota]
GWAP
Games with a Purpose [von Ahn et al]
Kohavi A/B
A/B Testing [Kohavi et al]
A B
Many-Eyes / sense.us
Many Eyes [IBM]
Mankoff Green FB
StepGreen [Mankoff et al]
RESEARCH
Research Approaches
Studying characteristics of online communities Collect usage data; Observe, interview users
Intervene in existing systems e.g., Facebook apps Controlled experimentation
Introduce + study new systems Requires massive investment (?)
Mining social media Recommendation and matching algorithms, …
Research Questions
How and why do people join communities?How is collective action organized?Why do people contribute?Issues of quality control, privacy, trust, …What are the interactions between social
structure and system design?
How do these findings generalize and inform the design of new socio-technical systems?
friendster
Profiles, Fakesters, Fraudsters
vizster
[InfoVis 05]
World of Warcraft
World of Warcraft [Yee, Ducheneaut et al]
History Flow
Wikipedia History Flow [Viégas et al]
“Talk” Pages on Wikipedia
Viégas et al. 2007
Coordination on Wikipedia
“… [we] note that administrative and coordinating elements seem to be growing at a faster pace than the bulk of articles in the encyclopedia [Wikipedia]”
Viégas et al. 2007
“Emergent” Order and Coordination
Wiki Dashboard
Wiki Dashboard [Suh et al]
Collaborative Tagging & Rating
[Chi & Mytkowicz]
[Budiu, Pirolli, & Hong]
Public Goods
Non-Excludability No one can be stopped from “using” the
good
Non-Rival Goods (Jointness of Supply) Consumption does not reduce availability
Commons-based peer-production (Tech-mediated) social production of goods
Free-rider problem Consumption disproportionate to production Gnutella: 10% users 87% of music [Adar ‘00]
Integrating ContributionsReduce the cost of synthesizing
contributions
Wikipedia: Shared RevisionsNASA ClickWorkers: Statistics
Incentives for Contribution [Benkler]
Monetary (you get reward, $) e.g.: Mechanical Turk
Hedonic (you enjoy it) e.g.: Games with a Purpose
Social-Psychological (you get social capital) e.g.: Discussion forums, Open-source
software
Manipulating Incentives
How to design systems to foster contribution?
Ling et al ‘05: movie recommendations Highlight unique contributions, issue
challenges
Cheshire ’07: simulated music sharing Positive feedback from peers: sustained
contrib. Visible group activity: short-lived boost
Social Psychological Incentives
1 2
block
0.50
0.60
0.70
0.80
0.90
Esti
mate
d M
arg
ina
l M
ean
s
Experiment Condition
High Social Approval
Low Social Approval
High Observational Cooperation
Low Observational Cooperation
Control Condition
[Cheshire]
Awareness and Task AllocationAn understanding of the activities of
others, which provides a context for your own activity. [Dourish & Belotti ‘92]
Ensure work is relevant to the group’s activity View the activities of others (e.g., live or via history) Coordination via shared artifacts Info explicitly generated or passively collected?
Scented Widgets [Willet et al, InfoVis 07]
Visual navigation cues embedded in interface widgets
Next Time… Research
The Science of Design, in The Sciences of the Artificial, pp. 128-159.Herbert A. Simon
The Structure of Scientific Revolutions, Ch. 1-4, 9, Thomas S. Kuhn, 1962, pp. 1-42, 92-110.Thomas S. Kuhn