The Design of Convivial Multiagent Systems

2002

Abstract In this paper, we present an abstract structure called dependence graph, an extension of the notion of dependence network, as proposed in [16]. While this latter can be applied to express a set of dependence relations of a single agent, this new structure can be applied to the multi-agent case. It can be used, therefore, for the study of emerging social structures, such as groups and collectives, and may form a knowledge base for managing complexity in both competitive and organisational or other cooperative contexts.

Proceeding of the 2008 conference on STAIRS 2008: …, 2008

Social networks are structures that aim to represent the relationships among the actors of a society. In the multiagent model, these networks depict the dependencies among the agents. The dependencies reflect the relation between the goals of the agents and the agents who have the power to achieve them. Like any social structure, also a multiagent system can be regulated by a set of norms and institutional powers that form an institution. Differently than informal social networks, the institutional social structure has an inherent dynamic character which cannot be captured by the current dependence networks. The networks have to reflect the changes of dependencies among agents created by the exercise of institutional powers. The aim of this paper is to define more dynamic social networks. From a description of an agent system it is possible to build a static dependence network. In the same way we describe how to pass from an abstract representation of what is an institution with its institutional powers to a dynamic social dependence network.

Abstract A Multi-Agent System (MAS) is an organization of coordinated autonomous agents that interact in order to achieve common goals. Considering real world social organizations as an analogy, this paper proposes architectural styles and design patterns for MAS which adopt concepts from social theories. The styles are intended to represent a macro-level architecture of a MAS in terms of actor, goal and actor dependency and are evaluated with respect to software quality attributes.

NORMAS, 2008

Social networks are structures that aim to represent the relationships among the actors of a society. In the multiagent model, these networks depict the dependencies among the agents. The dependencies reflect the relation between the goals of agents and agents who have the power to achieve them. Like any social structure, also a multiagent system have to be regulated by a set of norms and institutional powers that form an institution. Differently than informal social networks, the institutional social structure has an inherent dynamic character which cannot be captured by the current dependence networks. The networks have to reflect the changes of dependencies among agents created by the exercise of institutional powers. The aim of this paper is to define more dynamic social networks. From a description of an agent system it is possible to build a static dependence network. In the same way we describe how to pass from an abstract representation of what is an institution with its institutional powers to a dynamic social dependence network.

2005

This special issue contains four selected and revised papers from the second international workshop on normative multiagent systems, for short NorMAS07 (Boella et al. (eds) Normative multiagent systems. Dagstuhl seminar proceedings 07122, 2007), held at Schloss Dagstuhl, Germany, in March 2007. At the workshop a shift was identified in the research community from a legal to an interactionist view on normative multiagent systems.

2012

With the pervasive development of sociotechnical systems, such as Facebook, Twitter and digital cities, modelling and reasoning on social settings has acquired great significance. Hence, an independent soft objective of system design is to facilitate interactions. Conviviality has been introduced as a social science concept for multiagent systems to highlight soft qualitative requirements like user friendliness of systems. Roughly, more opportunity to work with other people increases the conviviality. In this paper, the question we address is how to design systems to increase conviviality by design. To evaluate conviviality, we model agent interactions using dependence networks, and define measures that quantify interdependence over time. To illustrate our approach we use a gaming example. Though, our methods can be applied similarly to any type of agent systems, which involve human or artificial agents cooperating to achieve their goals.

1999

Introduction Agents operate and exist in some environment, which typically is both computational and physical. The environment might be open or closed, and it might or might not contain other agents. Although there are situations where an agent can operate usefully by itself, the increasing interconnection and networking of computers is making such situations rare, and in the usual state

Abstract. In this paper, we consider the design of convivial multi-agent systems. Conviviality has recently been proposed as a social concept to develop multi-agent systems.

Artificial Intelligence Review, 2012

Dependence between heterogeneous agents has gained attention from researchers because dependence influences the capabilities of a particular agent in a multi-agent system. In this paper, we have reviewed and discussed the dependence graph which is also known as interdependence graph in some author's work. Dependence graph is originally proposed in the social reasoning mechanism to visualize the dependence network. In this paper, dependence network is explained and an in depth analysis of the dependence graph is presented as well. Next, we are going to discuss about the social phenomena and group formation of the dependence graph. In addition, the further works of other authors are listed in this paper. Also, we discuss the drawbacks and propose some potential solution for enhancing the dependence graph.

Journal of Telecommunication, Electronic and Computer Engineering, 2017

Distributed systems for decision support, crisis management, environment monitoring, e-government and smart cities, and for other areas, as well as Internet-of-Things applications are increasingly complex sociotechnical systems. Because of their distributed nature and complexity, such systems are cognitively hard to grasp and design. Therefore, appropriate simplified representations of complex systems – models – are required for designing and analyzing them. Moreover, a complex system needs to be simulated and prototyped before the development process can start to make sure a system with appropriate behavior has been designed. For modeling complex systems, “agent” or “actor” is a useful notion because conceptually sociotechnical systems consist of interacting human agents and man-made agents, such as smart software components and intelligent devices. In simulation and rapid prototyping, software agents can emulate both humans and technical components of a complex sociotechnical syst...