SEDNAM 2014 - International Workshop on Socio-Economic Dynamics: Networks and Agent-based Models

Investigation of socio-economic models faces three levels of difficulty: the agents definition, the choice of their interactions and the derivation of the associated macroscopic evolution from the chosen microscopic dynamics. Within such a framework, agent-based modeling combined with complex networks theory have opened the field of complex adaptive systems, where both the system's and the agents' structure co-evolve in a continuous interplay. Agents are thus allowed to locally interact with each other via a network of connections allowing both the agents' state (e.g., people's opinion in a social system or banks' liquidity in a financial system) and the agents' interactions (number and intensity) to eventually change in response to the neighbors' output, giving origin to a (time-discrete or time-continuous) dynamical process. As a result, some statistical regularities emerge, not derivable from the singular, microscopic behavior only: Sociophysics and Econophysics provide many examples, as shared opinions, cultures, languages, economic crises, bubbles of commodity prices, etc. Keeping up with statistical physics leads to interpret the emergent properties in terms of phase transitions (the ordered phase representing consensus and agreement, described by a low social temperature), diffusive processes (like the evolution of the price of stocks or options), turbulent phenomena (as the fluctuations in exchange rates between foreign currencies), magnetization (as the difference between the density of the agents having a different opinion) and so on. Such a modeling of socio-economic systems contrasts substantially with the traditional approach resting upon the axiomatic representative agent paradigm, according to which all agents in the systems are assumed to be isolated and to act in the same way, i.e., trying to optimize an utility function appropriately defined. On the contrary, modern agents-based models, often embedded into a complex network of interactions, allow to describe the dynamics of systems composed by strongly heterogeneous agents, characterized by particular behaviors (e.g. conformist, non-conformist, committed, etc.). Remarkably, agent-based models can be also embedded into multiple networks (multi-layer and multiplexes), with the aim of representing more realistic scenarios, e.g., interacting people on different online and offline social networks, interacting countries exchanging different kinds of commodities and more. The proposed satellite will be focused on the several ways to overcome the paradigm of the representative agent in socio-economic systems, by substituting the "intellectually-driven" axioms of traditional social and economic sciences with the data-driven analyses of statistical physics, within the frameworks of agent-based modeling and network theory. The test bench of the latter will precisely be the analysis of those "big data" that social and economic sciences are nowadays facing.
AIM OF THE SATELLITE
The aim of this satellite is to gather scholars, belonging to different fields such as physics, computer science, economics, sociology and linguistics, to review the state of the art of agent-based (and multi-agent) modeling, complex networks theory, complex adaptive systems and statistical physical modeling, which go beyond the representative agent paradigm. The relevance of the satellite to the conference's lines (in particular the 5th and 6th ones, Language, Linguistics Cognition and Social Systems and Economics and Finance) lies in the interdisciplinary trait of the proposed techniques, representing the core of the analyses carried on in the many branches of complex systems science. Scholars and researchers will be invited to submit contributions mainly (but not exclusively) on the following topics:
Agent-based modeling for socio-economic complex systems;
Spreading phenomena on socio-economic networks;
Statistical Physics models for socio-economic complex systems;
Multi-level socio-economic complex systems;
Big-data analysis of socio-economic systems;
Opinion and consensus dynamics;
Emergent phenomena in language dynamics;
Social networks and language evolution;
Systemic risk estimation for economic and financial networks.
CALL FOR ABSTRACTS
The satellite will be a half-day long and will host around 8-12 contributed talks. Submissions will be evaluated by the Scientific Committee members, based on the adherence with the themes of the satellite, their novelty and soundness. Abstracts are submitted via the EasyChair website. Authors who do not have an EasyChair account should sign up for an account (for identification purposes, make sure to use the same email address as the one used for the conference registration). Submissions are required to be at most two pages long including the following informations: title, author(s), affiliation(s), e-mail address(es), abstract (with a few figures). As the selection process will be completed, the contributors will be notified by e-mail.