A Multi-disciplinary Formal Approach to Mechanism Design for Peer Governance of Social Systems

12^th June 2012, 16:00 G12
Maryam Ghaffari Saadat
Department of Computer Science
The Graduate Center,
The City University of New York,
USA

Abstract

Over the past decade, much attention has been drawn to the application of peer governance in social, economic, and political settings. Peer governance is "a bottom-up mode of participative decision making" in which the outcome for each individual emerges from her interactions with other agents (horizontal power structure) as opposed to being decided by a subset of agents who are authorized to micromanage others (vertical power structure). In the absence of a central authority to closely supervise the behaviour of peer-to-peer social systems and (supposedly) intervene in the interest of the collective welfare, the performance of such systems is highly dependent on the systematic effectiveness of the interaction protocols in coordinating the individuals toward collectively beneficial arrangements. Thus peer governance could be described as management by design. This renders formal approaches an excellent toolbox for investigating whether such systems meet the required specifications.
In this talk I will propose interesting directions to cross-fertilise logic with game theory and social network analysis in order to develop a meta-framework for analysing the paradox of rationality and cooperation in human societies. To this end, a temporal extension of Social Network Logic will be presented to analyse a network-based coordination game of recycling and demonstrate the high capacity of such hybrid frameworks to model the emergence of norms and conventions. Furthermore, different allocation procedures developed for a two-player cake-cutting game will be evaluated using Alternating-time Temporal Logic to highlight the capacity of this framework to model strategic behaviour while capturing subtle, yet very influential, procedural differences which are often disregarded in game theoretical analysis. The talk closes with an outline of the proposed framework for modelling and analysing strategic behaviour in social networks.