Difference between revisions of "Schweitzer abs"

From Santa Fe Institute Events Wiki

Line 8: Line 8:
===Thursday, January 10, 2008===
===Thursday, January 10, 2008===
3:10 - 3:50 '''Frank Schweitzer''' ([ homepage])
3:10 - 3:50 '''Frank Schweitzer''' ([[Media:Sfi-talk-schweitzer-handout.pdf|slides]], [ homepage])
''Enhancing social interaction: preferences, similarities, and trust''
''Enhancing social interaction: preferences, similarities, and trust''

Latest revision as of 00:09, 11 January 2008

Workshop Navigation

Is There a Physics of Society? January 10-12, 2008, Santa Fe NM

Organizers: Michelle Girvan (University of Maryland) and Aaron Clauset (Santa Fe Institute)

Thursday, January 10, 2008

3:10 - 3:50 Frank Schweitzer (slides, homepage)

Enhancing social interaction: preferences, similarities, and trust

Sociophysics - as any other kind of scientific enterprise - undergoes a three-step evolution: (i) empirics -- all the power laws are derived from social (network) data, (ii) modeling -- KISS (keep it simple and stupid) type agents interact in a reductionistic manner to reproduce the stylized facts found under (i), and (iii) application -- here we may ask ourselves what we have actually achieved during steps (i) and (ii), and how we can make use of it all, if any. Interestingly, not all scientific enterprise make it to step (iii).

I argue that steps (i) and (ii) of sociophysics may be used as an input for designing social interaction conducted by artifical devices, such as technical gadgets or Web 2.0 technologies. Social interaction occurs between real people, but specific algorithms may be designed to improve this interaction in a virtual space in such a way that the overall performance of the social system is improved. Successful examples of such algorithms already exist - e.g. in recommender systems for online shops or dating websites. However, they still have to overcome problems known from other kind of social interaction, such as trustworthiness, free riding and malicious attacs. As I will demonstrate, this can be solved by combining such algorithms with existing social networks.