8/28/2009
Social Computing: a New Interdisciplinary Study Julita Vassileva Computer Science Department University of Saskatchewan 1
What is Social Computing? •
•
Social computing is a social structure in which technology puts power in communities, not institutions. As more individuals use the Internet to shop, work, and exchange ideas, a more egalitarian social structure is emerging. Individuals g g g g take cues from one another, rather than traditional sources of authority — like corporations, media outlets, political institutions or organized religions. Manifestations of social computing include:
• • • • • • • • • • • • • • •
Social networks Peer‐to‐peer content distribution Open source software Blogs RSS Podcasting Consumer‐to‐consumer commerce Meet‐ups Mash‐ups Tagging Social search User‐generated content Peer ratings Wikis Comments and trackbacks Widgets
• •
Voter‐driven content (Forrester Research, 2008) http://www.forrester.com/ResearchThemes/SocialComputing
Key "tenets of social computing" outlined by Charlene Li: •innovation will shift from top‐down to bottom‐up •value will shift from ownership to experience l ill hift f hi t i •power will shift from institutions to communities •http://www.socialcustomer.com/2006/02/the_forrester_s.html
2/25
1
8/28/2009
Complex Systems Sociology, Anthropology h l
Computer Science, Web Social Computing
Decision Making, Politics, Education
Social Psychology Behavioral Economics
3/25
Computer Science • Social Computing evolved as a way of i t interacting and collaborating on the web ti d ll b ti th b
4/25
2
8/28/2009
Social Sciences • Analyzing the interactions in communities • Observing social phenomena – hazing of newbies in forums (e.g. X‐Files fans) C. Honeycutt (2005) Hazing as a Process of Boundary Maintenance in an Online Community
– reputation /power economy of Wikipedia (similar to that of research community) A.Forte, A.Bruckman (2005) Why do people write for Wikipedia? Georgia Tech Report
5/25
Behavioral Economics • Why do people behave irrationally / altruistically? lt i ti ll ? • Money‐economy vs. social norms – E.g. try to pay your mother‐in‐law for the lovely Thanksgiving dinner she cooked for the family – Reciprocation (immediate, delayed, concrete, p ( , y , , generalized) – Gift economies Dan Ariely (2007) Predictably Irrational 6 / 25
3
8/28/2009
Social Psychology • Individual motivations for contribution – Many theories can explain observed behavior – Can a theory be used as a guideline in system design to ensure motivation? Rob Kraut (2005) Social Psychology & Online communities
– Exploring the effect of visualization according to Exploring the effect of visualization according to certain theories in different communities • Social comparison theory in Comtella • Common identity theory in WISETales • Common bond theory 7/25
Incentive: Status/Reputation Customer Loyalty Programs
Image from depts.washington.edu/.../painting/4reveldt.htm Cheng R., Vassileva J. (2006) Design and Evaluation of an Adaptive Incentive Mechanism for Sustained Educational Online Communities. User Modelling and User-Adapted Interaction, 16 (2/3), 321-348.
8/25
4
8/28/2009
9
Immediate gratification for rating
Topics and individual postings that are rated higher appear “hot”, those rated lower appear “cold” Æ colours ease navigation in the content Æ aesthetically pleasing, intuitive
Webster A.S., Vassileva J. (2006) Visualizing Personal Relations in Online Communities, Proceedings Adaptive Hypermedia and Adaptive Web-Based Systems, Dublin, Springer LNCS 4018, 223-233. 10/25
5
8/28/2009
Sahib, Z., Vassileva J. (2009) Designing to Attract Participation In A Niche Community For Women In Science & Engineering, in Proc.WS Social Computing in Education, with the 1st IEEE International Conference on Social Computing, SocialComp'2009, Vancouver, BC, August 29-31, 2009.
11/25
Common bond ‐ reciprocation
Raghavun, K., Vassileva J. (2009) Visualizing Reciprocal and non-Reciprocal Relationships in an Online Community. Proc. Workshop on Adaptation and Personalization for Web 2.0, in conjunction with UMAP 2009, June 22-26, 2009, Trento, Italy. 12/25
6
8/28/2009
Business/Organizational Studies • How do groups make decisions? • Features of groups that make good decisions: diversity, decentralization, independence, aggregation • Phenomena: cascades, social norms, group think, • Interactions: fairness, punishment, trust
Cass Sunstein (2007) Infotopia James Surowiecki (2007) The Wisdom of Crowds 13/25
How are small groups different from wise crowds? • People think of themselves as members of a team, while in a market, they think of themselves as independent actors. • The group has an identity of its own – Consensus is important for the existence and comfort of the group – Influence of the people in the group on each other’s judgment is unavoidable. – Group polarization Group polarization
• Collective wisdom, in contrast, is something that emerges as a result of many different independent judgments, not something that the group should consciously come up with. 14/25
7
8/28/2009
Consequences • Small cohesive groups / communities may be wrong or biased (encapsulation) • Does this apply to online groups ? • Currently we see tagging, voting (rating) systems and recommenders emerge as forms of “collective wisdom” online
• O Open question: what can designers do to i h d i d avoid biases resulting from activities of small groups online? 15/25
Importance of mechanism • A decentralized system can only produce intelligent results if there is a means of aggregating the private information of there is a means of aggregating the private information of everyone • An aggregation mechanism is a form of centralization, (ideally) of all the private information of the participants – provides incentives for revealing truthfully private info – should not inject extra bias in the system Mechanisms: New mechanisms: – One person with foresight - Prediction markets - Trust and reputation – Deliberation mechanisms – Polls / votes – Price in a open market 16/25
8
8/28/2009
Complex, self organizing systems
N(k) ‐ # pages with K incoming links N(k) ~ k –γ , where γ – degree exponent, in this case γ = 2.5
Many empirically observed networks appear to be scale-free: world wide web, protein networks, citation networks, and some social networks. 17/25
Scale Free Networks • Macroscopic effects of individual behaviour – emerging patterns (Barabási & Albert, 1999) – Growth and preferential attachment explain the hubs and power laws in complex networks, like the Web;
• Fitness of a node in a competitive environment • The “Fit get rich” model (borrowing formalisms from quantum mechanics) predicts a phenomenon called Einstein‐Bose condensation • In some networks (under special conditions) all links will ultimately point to one node: “The winner takes it all” or
18/25
9
8/28/2009
Robust Scale Free Networks • Scale‐free networks are extremely robust in case of random failures random failures • Studying network resilience – In random networks, some node failures can easily break a network into isolated, non‐communicating parts. – Yet, a study of the Internet resilience showed that we can remove 80% of all nodes, and the remaining 20% will still , g remain connected – The key to this is the presence of hubs, removing nodes randomly is not likely to affect them, and they hold the NW together 19/25
Vulnerable Scale Free Networks • Yet, scale‐free NW are very vulnerable to g g targeted attacks and to cascading failures • In case of targeted attack on a critical number of hubs, the network disintegrates very quickly • Cascading failures – examples – Power grid black outs (1996, 2003) – Cascades of malfunctioning routers on the Internet – Cascading East Asian economic crisis in 1997 – Cascades in ecological habitats 20/25
10
8/28/2009
Consequences • The laws of power networks lead to concentration – clear targets that need to be protected – less diversity (or lesser impact of diverse opinion), less creativity – more power (network power, $$$s, legal advisors and lobbyists) in very few hands – possibility of possibility of “locking locking up up” the internet by a couple of the internet by a couple of corporate giants • Creeping copyright protections (patents, DRM) • Apple locking up the iPhone 21/25
Spreading Viruses and Innovation • Viruses • Innovation
# adopters
• Hubs: – – – –
Opinion leaders time Power users Laggards Innovators Hubs Mass Influencers Are not necessarily innovators, but they are key to spreading y , y y p g an innovation, launching an idea….
• Yet, not all innovations catch on (e.g. Apple’s Newton). Why some do and some do not? • Diffusion models 22/25
11
8/28/2009
Disease diffusion models • Threshold model: Each innovation has – spreading rate – the likelihood that it will be adopted by a person introduced to it, and introduced to it, and – critical threshold – defined by the properties of the NW in which the information spreads – If spreading rate < critical threshold, it will die, Else, the number of people adopting the innovation will increase exponentially.
• This model has been used by epidemiologists, marketers, sociologists, political scientists – but it doesn’t explain the persistence of some viruses like AIDS – It assumes a random network topology. – In scale‐free topology, the critical threshold disappears. 23/25
Consequences • Ideas can be spread very quickly and far in a scale free network l f t k • Political ideas, innovations, but also radical / extremist ideas • Action can be organized very quickly – E.g. E g “flash flash‐crowds crowds” with Twitter with Twitter
• Are we prepared to deal with this? • What is the impact on education? 24/25
12
8/28/2009
Some food for thought… “While entirely of human design, the Internet now lives a life of its own. It has all the characteristics of a f f f complex evolving system, making it more similar to a cell than a computer chip. Many diverse components, developed separately, contribute to the functioning of a system that is far more than the sum of its parts. Therefore Internet researchers are increasingly morphing from designers into explorers. They are like bi l i biologists or ecologists who are faced with an l i h f d ih incredibly complex system that, for all practical purposes, exists independently of them.” (pp.149‐150) Albert‐László Barabási, Linked, Plume Publ. 2003. 25/25
13