Description of work
A generic property of social, biological, political and economic networks is their ability to evolve in time, creating and suppressing interactions, including self-destructive patterns. Even though there are external field influences, and perturbations which cannot be avoided, the systems usually self-organize in such a way that the structure of links in the network is quite inhomogeneous, revealing the occurrence of cooperator “leaders” with a very high connectivity, which guarantee that global cooperation can be sustained in the whole network. A model of “connected scientists” will take into account that the individual decisions of every researcher lead to cooperative changes in science development. These individual decisions will be modeled by local social fields acting on the scientist from his/her scientific environment (literature, personal contacts, influence of scientific authorities, economic conditions). As the result of the changing environment the active agent representing a scientist will undergo discontinuous changes of its internal variables (scientific interests). The creation of this model will form task T2.1.
We approach these issues within the framework of an adaptive network of unusually linked agents, - in order to get some idea of the control parameters, and hence to emphasize (or reduce) the crisis strengths, frequencies, and even forecast their timing. It is fundamental to realize that interactions between agents should be based on both "energy/cost considerations" and also on "entropy/information content". It should be stressed that since the systems are (usually) out of equilibrium, the crises are to be described through statistical concepts, such as probability distribution functions, with their intrinsic evolution equations, the master (or Fokker-Planck) equation, and the subsequent Langevin equation.
Conditions for triggering scientific avalanches corresponding to the control of crises in active agent networks will be investigated. Here results of stochastic resonance in network systems will be used. The results should help in T2.2 toward describing networks, hence suggesting models with specific control parameters for systems with evolving groups, population distributions, and hierarchy formations in social and scientific networks, but also other non equilibrium dynamic phenomena (involving birth and death processes) prone to crises and unusually strong events.
In fact, it has already been argued, on a more simple model, based on the sand pile analogy that a non-trivial feedback loop between the volatility and the variance (of the ''agent wealth'') leads to non-trivial patterns in the overall (in the case of financial market) trends. If the feedback is too strong, it may even endanger the stability. Hence the generalization to crises in active social networks is of great interest. This study has shown that the control of crises has analogies with the control of sand pile avalanches.
Data acquisition on specific topics from other WP's will serve to select the parameters, and hence models, from the most simple to the more complex for the simulation tasks, - moreover suggesting experimental (field) investigations and feedback through forecast verifications.
T2.1. Crises in networks of agents simulating discontinuous science evolution.
T2.2. Control of crises in agents networks |
