The main objective of this project is to understand and develop design strategies able to cope with the complex 2-way dynamics of sociality, consisting of emergent and immmergent processes: from interaction among individual agents to aggregate level, and immergence of entities (norms) at the aggregate level into agents' minds. In particular, we plan to focus on norm innovation. As research priorities, besides dealing with incompleteness and uncertainty, we intend to contribute to the understanding and description of hierarchic systems by describing agents acting on multiple, i.e. individual, communitarian and institutional levels. As to understanding of distributed processes in IT, the project is aimed at modeling the interactive, bi-directional processes of emergence. A summary of main theoretical goals is: understand and manage complexity in social systems with autonomous agents; understand how new conventions and norms emerge and spread in these systems; study of norm innovation by means of agent-based simulation.

The main technological aim of the project is to construct a simulator for exploring and experimenting upon norm-innovation. Concerning the applicative side, we intend to contribute to the regulation of e-communities by handing out a simulator for the emergence of new norms in complex social systems, where situated experiments can be run. While the simulator will be designed as a general-purpose tool, a specific study case will be selected as so to provide the necessary grounding parameters.

 

 

Complex systems are defined as systems with many interdependent parts which give rise to non-linear and emergent properties determining the high-level functioning and behavior of such systems. Due to the interdependence of their constituent elements and other characteristics of complex systems, it is difficult to predict system behavior based on the 'sum of their parts' alone. Examples of complex systems include bee hives, bees themselves, human economies and societies, nervous systems, molecular interactions, cells and living things, ecosystems, as well as modern energy or telecommunication infrastructures. Arguably one of the most striking properties of complex systems is that conventional experimental and engineering approaches are inadequate to capture and predict the behavior of such systems.

 To complement the conventional experimental and engineering approaches, computer-based simulations of complex natural phenomena and complex man-made artifacts are increasingly employed across a wide range of sectors. Typically, such simulations require computing environments which meet very high specifications in terms of processing units, primary and secondary storage, and communication. Supercomputers constitute the de facto technology to deliver the required specifications. Acquiring, operating and maintaining supercomputers involve considerable costs, which many organizations cannot afford. The working assumption of the QosCosGrid project is that a grid could be enhanced by suitable middleware to provide features and performance characteristics that resemble those of a supercomputer. We refer to such a grid as quasi-opportunistic supercomputer. The QosCosGrid projects aim is to develop such a system.

 

 

Traffix is a general framework for agent-based traffic simulation built on the Repast 3.1 agent-based simulation platform. This framework is suitable for the modelling and simulation of a wide range of traffic scenarios, ranging from individual intersections or interchanges, to a city-wide traffic network. The framework enables modelers with rudimentary Java skills, to build complex traffic scenarios, to specify different driver behaviors and cars with different characteristics. This allows for experimentation with various composition of the drivers' population with regard to lane changing, car following, or route planning within the same overall traffic scenario.

 

 

This R&D project is sponsored by the Hungarian government. Its overarching goal is to make the application of telephony available to users with hearing or speaking difficulties. Exploiting our expertise in adaptive agent technologies, the task of our group is to develop an adaptive dialogue interface to help users with speaking challenges to enter text to be converted to speech and transferred over the telephony line. The main focus of this interface is to learn the user's way of speaking and offer relevant lines and sequences depending on the particular dialogue and on the actual position within it.

 

 

This project addresses the problem of generating networks that are robust against random failures (i.e., against the random removal of nodes). More precisely, we are considering the generation of networks with skewed degree distributions. We construct a family of agent-based model in which agents represent the nodes of the network that connect to one another aiming to maximize their connectivity. Each agent can build a fixed number of links.

However, information about the existing network is costly, so the agents must optimize under budget constraints, i.e., only having information about a limited number of existing nodes. The model is related to the model of Simon that has become more widely known in its variant by Albert and Barabási, the 'preferential attachment' model. The main difference between these models and those investigated within this project is that the former models require global access to information (i.e., the arriving new agent or node has to assess the distribution of links/size of the whole pre-existing population), while our models operate with limited information and utility  maximization subject to this set of information. Access to information is governed by economic means in our models.

 

 

Trust and cooperation among agents in a multi-agent system have several interpretations, including shared preferences, work toward a common goal, dependable expertise and opinions, altruistic behavior, etc.. Yet, in most fields of the social sciences, trust is understood in the strategic context: i.e., as the (missing) tendency of the agent to take unilateral advantage over its partners. The classic framework to explore such situations is the (Iterated)  Prisoner's Dilemma game (IPD). Following the series of works by Axlerod, this framwork is now  ubiquotious in many disciplines and forms the base of many models, studying a variety of problems.

The role of interaction topology (i.e., network structure) was also studied in IPD settings. In their seminal paper, Cohen, Axelrod and Riolo (CAR) found that, in their evolutionary model, context-preservation (i.e., a static network) and not network structure was the key for the survival (and domination) of altrustic/cooperative behavior (i.e., agents playing TFT). When network dynamics was allowed, ALLD agents took over the world. While this result has important and appealing interpretations in social disciplines, it is in stark contrast with what is known about information/behavior cascades and adaptation in complex (social) networks. In the latter field, it is established that the average path-length of the network plays an important role in the spreading of information and behavior.

This theoretical project investigates network dependence in IPD games with the aim to consolidate the previously discussed apparent contradiction. It departs from classic models by assuming individual learning, instead of evolutionary adaptation. The study of this altered model on various networks shows that context-preservation is not always sufficient for achieving a cooperative regime. Furthermore, and more importantly, the results suggest that the performance of the TFT strategy may depend on network structure and the speed of learning.

 

 

The direct route to artifacts is via design, i.e. by specification and subsequent realization.

Evolutionary Technology looks for an indirect route. The task is to produce an abundance of forms and functions of a practically endless variety by means of evolutionary methods. This implies the twin challenges of the 'arrow of complexity' and 'open-ended evolution', i.e. of producing increasingly complex machineries in the course of time, and doing that in a persistent, self-supporting process propelled by entirely endogenous causes.

 

Open-ended evolution is widely recognized as a difficult and unsolved problem. John Holland, the founder of Genetic Algorithms has recently asked: "Can we produce an existence-proof model, akin to von Neumann's model of self-reproduction, that exhibits open-ended evolution, with increasing diversity and complexity? ". Also, the 'arrow of complexity' hypothesis is challenged from several directions. Currently there is no accepted general evolutionary theory for the origin of complexity or the maintenance of evolutionary change.

 

In the above terms, we understand Evolutionary Technology as an attempt to realize open-ended evolution showing complexification and to harness it for technological needs. The FATINT project deals with the first part of this problem (i.e. the growing space of functionalities and attributes), leaving issues of controllability aside.

 

 

 

This project dealt with the application of data mining techniques to study extremely large connection networks and group detection in telecommunications data warehouses. The task of our group was to desing and implement innovative data visualization techniques, methods and applications to enable interactive analysis. In addition, we were responsible for integrating the final user application.