In this paper we propose a new role for the agent metaphor in the definition of the organizationa... more In this paper we propose a new role for the agent metaphor in the definition of the organizational structure of multiagent systems. The agent metaphor is extended to consider as agents also social entities like organizations, groups and normative systems, so that mental attitudes can be attributed to them -beliefs, desires and goals -and also an autonomous and proactive behavior. We show how the metaphor can be applied also to structure organizations in functional areas and roles, which are described as agents too. Thus, the agent metaphor can play a role similar to the object oriented metaphor which allows structuring objects in component objects. Finally, we discuss how the agent metaphor addresses the problems of control and communication in such structured organizations.
Proceedings of the 6th international conference on Electronic commerce - ICEC '04, 2004
In this paper we address the problem of defining roles in organizations like e-trade ones. The me... more In this paper we address the problem of defining roles in organizations like e-trade ones. The methodology we use is to model roles according to the agent metaphor: we attribute to roles mental attitudes, like beliefs, desires and goals, we relate them to the agent's required expertise and responsibilities, and we model role behavior in game theoretic terms. Analogously, the organization is modelled as an agent which acts as a normative system: it imposes obligations to roles and to the agents playing the roles.
International Journal of Cooperative Information Systems, 2007
Normative multiagent systems provide agents with abilities to autonomously devise societies and o... more Normative multiagent systems provide agents with abilities to autonomously devise societies and or- ganizations coordinating their behavior via social norms and laws. In this paper we study how agents negotiate new social norms and when they accept them. We introduce a negotiation model based on what we call the social delegation cycle, which explains the negotiation of new social norms from agent desires in three steps. First individual agents or their representatives negotiate social goals, then a social goal is negotiated in a social norm, and finally the social norm is accepted by the agents when it leads to fulfilment of the desires the cycle started with. We characterize the allowed propos- als during social goal negotiation as mergers of the individual agent desires, and we characterize the allowed proposals during norm negotiation as both joint plans to achieve the social goal (obligations associated with the norm) and the associated sanctions or rewards (a control system associated with the norm). The norm is accepted when the norm is stable in the sense that agents will act according to the norm, and effective in the sense that fulfilment of the norm leads to achievement of the agents' de- sires. We also compare norm negotiation with contract negotiation and negotiation of the distribution of obligations.
Multiagent Systems, Artificial Societies, and Simulated Organizations, 2003
Even if, at first sight, deciding whether to respect obligations and coordinating cooperation see... more Even if, at first sight, deciding whether to respect obligations and coordinating cooperation seems two unrelated behaviors, they are both grounded in the social rationality of agents: in fact, they depend on the ability of predicting the actions of other agents. In this paper, we present a proposal for managing anticipatory coordination, which is used to regulate the autonomy of and to control agents in a definition of cooperation among agents and in a framework for dealing with obligations.
Normative sentences can be used to change or to describe the normative system, known as prescript... more Normative sentences can be used to change or to describe the normative system, known as prescriptive and descriptive obligations respectively. In applications of deontic logic it is important to distinguish these two uses of normative sentences. In this paper we show how they can be distinguished and analysed in a dynamic epistemic deontic logic.
Privacy policies are often defined in terms of permitted messages. Instead, in this paper we deri... more Privacy policies are often defined in terms of permitted messages. Instead, in this paper we derive dynamically the permitted messages from static privacy policies defined in terms of permitted and obligatory knowledge. With this new approach, we do not have to specify the permissions and prohibitions of all message combinations explicitly. To specify and reason about such privacy policies, we extend a multi-modal logic introduced by Cuppens and Demolombe with update operators modeling the dynamics of both knowledge and privacy policies. We show also how to determine the obligatory messages, how to express epistemic norms, and how to check whether a situation is compliant with respect to a privacy policy. We axiomatize and prove the decidability of our logic.
The development of complex distributed AI systems with heterogeneous and diverse knowledge is a c... more The development of complex distributed AI systems with heterogeneous and diverse knowledge is a challenge. System components must interact, coordinate and collaborate to manage scale and complexity of task environments targeting persistency and maybe, evolution of systems. Managing scale and complexity requires organized intelligence; in particular intelligence manifested in organizations of agents, by individual strategies or collective behaviour. System architects have to consider: the inter-operation of heterogeneously designed, developed or discovered components (agents, objects/artefacts, services provided in an open environment); inter-connection which cross legal, temporal, or organizational boundaries; the absence of global objects or centralised controllers; the possibility that components will not comply with the given specifications; and embedding in an environment which is likely to change, with possible impact on individual and collective objectives.
In this paper we describe the extension of the EasyDial developer interface of Dialogos, the spok... more In this paper we describe the extension of the EasyDial developer interface of Dialogos, the spoken dialog system of Loquendo. EasyDial has been integrated with an XML-based representation of data structures and procedural knowledge which are dependent on the application domain. A set of XSLT programs translates this knowledge into textual data and C language procedures which are integrated in Dialogos after the processing by EasyDial. The adoption of an XML-based representation for declaring the domain dependent knowledge speeds up the application development process significantly.
We study the design of policies for virtual communities of agents based on peer-to-peer systems o... more We study the design of policies for virtual communities of agents based on peer-to-peer systems or the grid infrastructure. In a virtual community agents can play both the role of resource consumers and the role of resource providers. Moreover, the agents remain in control of their resources, and therefore we distinguish between the authorization to access a resource given by the virtual community and the permission to do so issued by the resource providers. We propose a logical multiagent framework for virtual communities that distinguishes three roles: resource consumption, provision, as well as authorization.
In this paper we study how roles can be added to patterns modelling relationships in Object Orien... more In this paper we study how roles can be added to patterns modelling relationships in Object Oriented programming. Relationships can be introduced in programming languages either by reducing them to attributes of the objects which participate in the relationship, or by modelling the relationship itself as a class whose instances have the participants of the relationships among their attributes. However, even if roles have been recognized as an essential component of relationships, also in modelling languages like UML, they have not been introduced in Object Oriented programming when it is necessary to model relationships. Introducing roles allows to add attributes and behaviors to the participants in the relationship, rather than to the relationship itself, and to distinguish the natural types of the participants in the relationships from the roles the participants acquire in the relationships. We show how the role model of the language powerJava can be used to endow the relationship as attribute pattern with roles.
This paper begins with the comparison of the messagesending mechanism, for communication among ag... more This paper begins with the comparison of the messagesending mechanism, for communication among agents, and the methodinvocation mechanism, for communication among objects. Then, we describe an extension of the method-invocation mechanism by introducing the notion of "sender" of a message, "state" of the interaction and "protocol" using the notion of "role", as it has been introduced in the powerJava extension of Java. The use of roles in communication is shown by means of an example of protocol.
Social roles structure social institutions like organizations in Multi-Agent Systems (MAS). In th... more Social roles structure social institutions like organizations in Multi-Agent Systems (MAS). In this paper we describe how to introduce the notion of social role in programming languages. To avoid the commitment to a particular agent model, architecture or language, we decided to extend Java, the most prominent object oriented programming language, by adding social roles. The obtained language allows an easier implementation of MAS's w.r.t. the Java language. We also show that many important properties of social roles, studied in the MAS field, can be applied to objects. Two are the essential features of social roles according to an analysis reported in the paper: social roles are defined by other entities (called institutions), and when an agent plays a role it is endowed with powers by the institution that defines it. We interpret these two features into the object oriented paradigm as the fact that social roles are objects, which are defined in and exist only inside other objects (corresponding to institutions), and that, through a role, external objects playing the role can access to the object (institution) the role belongs to.
This paper is about control mechanisms for virtual organizations. As a case study, we discuss the... more This paper is about control mechanisms for virtual organizations. As a case study, we discuss the Renewables Obligation (RO), a control mechanism that was introduced in the United Kingdom to stimulate the production of renewable energy. We apply a conceptual model based on normative multiagent systems (NMAS). We propose to model both the participants and the normative system as autonomous agents, having beliefs and goals. Norms, which can be internalized by the agents as obligations, are translated into conditional beliefs and goals of the normative system, which concern both detection and sanctioning measures. We show that the model can handle both the regulative and the evidential aspects of the case.
Proceedings of the fifth international joint conference on Autonomous agents and multiagent systems - AAMAS '06, 2006
In this paper we illustrate how a role-based semantics for agent communication languages can embe... more In this paper we illustrate how a role-based semantics for agent communication languages can embed the two main existing models of agent communication languages, respectively based on 'mental attitudes' and 'social commitments' semantics. These two models have been presented as incompatible approaches, but recently we illustrated for persuasion dialogues and using our normative multiagent systems framework, that they can be seen also as complimentary ones. Independently from our own multi-agent model, in this paper we illustrate for the speech act 'inform' how the role based semantics embeds the other two semantics.
There are two main traditions in defining a semantics for agent communication languages, based ei... more There are two main traditions in defining a semantics for agent communication languages, based either on mental attitudes or on social commitments. In this paper, we translate both traditions in a different approach in which the dialogue state is represented by the beliefs and goals publicly attributed to the roles played by the dialogue participants. On the one hand, this approach avoids the problems of mentalistic semantics, such as the unverifiability of private mental states. On the other hand, it allows to reuse the logics and implementations developed for FIPA compliant approaches.
We present an algorithm for replanning in a reactive agent architecture which incorporates decisi... more We present an algorithm for replanning in a reactive agent architecture which incorporates decision-theoretic notions to drive the planning and metadeliberation process. The deliberation component relies on a refinement planner which produces plans with optimal expected utility. The replanning algorithm we propose exploits the planner's ability to provide an approximate evaluation of partial plans: it starts from a fully refined plan and makes it more partial until it finds a more partial plan which subsumes more promising refinements; at that point, the planning process is restarted from the current partial plan.
In this paper we propose a new role for the agent metaphor in the definition of the organizationa... more In this paper we propose a new role for the agent metaphor in the definition of the organizational structure of multiagent systems. The agent metaphor is extended to consider as agents also social entities like organizations, groups and normative systems, so that mental attitudes can be attributed to them -beliefs, desires and goals -and also an autonomous and proactive behavior. We show how the metaphor can be applied also to structure organizations in functional areas and roles, which are described as agents too. Thus, the agent metaphor can play a role similar to the object oriented metaphor which allows structuring objects in component objects. Finally, we discuss how the agent metaphor addresses the problems of control and communication in such structured organizations.
Proceedings of the 6th international conference on Electronic commerce - ICEC '04, 2004
In this paper we address the problem of defining roles in organizations like e-trade ones. The me... more In this paper we address the problem of defining roles in organizations like e-trade ones. The methodology we use is to model roles according to the agent metaphor: we attribute to roles mental attitudes, like beliefs, desires and goals, we relate them to the agent's required expertise and responsibilities, and we model role behavior in game theoretic terms. Analogously, the organization is modelled as an agent which acts as a normative system: it imposes obligations to roles and to the agents playing the roles.
International Journal of Cooperative Information Systems, 2007
Normative multiagent systems provide agents with abilities to autonomously devise societies and o... more Normative multiagent systems provide agents with abilities to autonomously devise societies and or- ganizations coordinating their behavior via social norms and laws. In this paper we study how agents negotiate new social norms and when they accept them. We introduce a negotiation model based on what we call the social delegation cycle, which explains the negotiation of new social norms from agent desires in three steps. First individual agents or their representatives negotiate social goals, then a social goal is negotiated in a social norm, and finally the social norm is accepted by the agents when it leads to fulfilment of the desires the cycle started with. We characterize the allowed propos- als during social goal negotiation as mergers of the individual agent desires, and we characterize the allowed proposals during norm negotiation as both joint plans to achieve the social goal (obligations associated with the norm) and the associated sanctions or rewards (a control system associated with the norm). The norm is accepted when the norm is stable in the sense that agents will act according to the norm, and effective in the sense that fulfilment of the norm leads to achievement of the agents' de- sires. We also compare norm negotiation with contract negotiation and negotiation of the distribution of obligations.
Multiagent Systems, Artificial Societies, and Simulated Organizations, 2003
Even if, at first sight, deciding whether to respect obligations and coordinating cooperation see... more Even if, at first sight, deciding whether to respect obligations and coordinating cooperation seems two unrelated behaviors, they are both grounded in the social rationality of agents: in fact, they depend on the ability of predicting the actions of other agents. In this paper, we present a proposal for managing anticipatory coordination, which is used to regulate the autonomy of and to control agents in a definition of cooperation among agents and in a framework for dealing with obligations.
Normative sentences can be used to change or to describe the normative system, known as prescript... more Normative sentences can be used to change or to describe the normative system, known as prescriptive and descriptive obligations respectively. In applications of deontic logic it is important to distinguish these two uses of normative sentences. In this paper we show how they can be distinguished and analysed in a dynamic epistemic deontic logic.
Privacy policies are often defined in terms of permitted messages. Instead, in this paper we deri... more Privacy policies are often defined in terms of permitted messages. Instead, in this paper we derive dynamically the permitted messages from static privacy policies defined in terms of permitted and obligatory knowledge. With this new approach, we do not have to specify the permissions and prohibitions of all message combinations explicitly. To specify and reason about such privacy policies, we extend a multi-modal logic introduced by Cuppens and Demolombe with update operators modeling the dynamics of both knowledge and privacy policies. We show also how to determine the obligatory messages, how to express epistemic norms, and how to check whether a situation is compliant with respect to a privacy policy. We axiomatize and prove the decidability of our logic.
The development of complex distributed AI systems with heterogeneous and diverse knowledge is a c... more The development of complex distributed AI systems with heterogeneous and diverse knowledge is a challenge. System components must interact, coordinate and collaborate to manage scale and complexity of task environments targeting persistency and maybe, evolution of systems. Managing scale and complexity requires organized intelligence; in particular intelligence manifested in organizations of agents, by individual strategies or collective behaviour. System architects have to consider: the inter-operation of heterogeneously designed, developed or discovered components (agents, objects/artefacts, services provided in an open environment); inter-connection which cross legal, temporal, or organizational boundaries; the absence of global objects or centralised controllers; the possibility that components will not comply with the given specifications; and embedding in an environment which is likely to change, with possible impact on individual and collective objectives.
In this paper we describe the extension of the EasyDial developer interface of Dialogos, the spok... more In this paper we describe the extension of the EasyDial developer interface of Dialogos, the spoken dialog system of Loquendo. EasyDial has been integrated with an XML-based representation of data structures and procedural knowledge which are dependent on the application domain. A set of XSLT programs translates this knowledge into textual data and C language procedures which are integrated in Dialogos after the processing by EasyDial. The adoption of an XML-based representation for declaring the domain dependent knowledge speeds up the application development process significantly.
We study the design of policies for virtual communities of agents based on peer-to-peer systems o... more We study the design of policies for virtual communities of agents based on peer-to-peer systems or the grid infrastructure. In a virtual community agents can play both the role of resource consumers and the role of resource providers. Moreover, the agents remain in control of their resources, and therefore we distinguish between the authorization to access a resource given by the virtual community and the permission to do so issued by the resource providers. We propose a logical multiagent framework for virtual communities that distinguishes three roles: resource consumption, provision, as well as authorization.
In this paper we study how roles can be added to patterns modelling relationships in Object Orien... more In this paper we study how roles can be added to patterns modelling relationships in Object Oriented programming. Relationships can be introduced in programming languages either by reducing them to attributes of the objects which participate in the relationship, or by modelling the relationship itself as a class whose instances have the participants of the relationships among their attributes. However, even if roles have been recognized as an essential component of relationships, also in modelling languages like UML, they have not been introduced in Object Oriented programming when it is necessary to model relationships. Introducing roles allows to add attributes and behaviors to the participants in the relationship, rather than to the relationship itself, and to distinguish the natural types of the participants in the relationships from the roles the participants acquire in the relationships. We show how the role model of the language powerJava can be used to endow the relationship as attribute pattern with roles.
This paper begins with the comparison of the messagesending mechanism, for communication among ag... more This paper begins with the comparison of the messagesending mechanism, for communication among agents, and the methodinvocation mechanism, for communication among objects. Then, we describe an extension of the method-invocation mechanism by introducing the notion of "sender" of a message, "state" of the interaction and "protocol" using the notion of "role", as it has been introduced in the powerJava extension of Java. The use of roles in communication is shown by means of an example of protocol.
Social roles structure social institutions like organizations in Multi-Agent Systems (MAS). In th... more Social roles structure social institutions like organizations in Multi-Agent Systems (MAS). In this paper we describe how to introduce the notion of social role in programming languages. To avoid the commitment to a particular agent model, architecture or language, we decided to extend Java, the most prominent object oriented programming language, by adding social roles. The obtained language allows an easier implementation of MAS's w.r.t. the Java language. We also show that many important properties of social roles, studied in the MAS field, can be applied to objects. Two are the essential features of social roles according to an analysis reported in the paper: social roles are defined by other entities (called institutions), and when an agent plays a role it is endowed with powers by the institution that defines it. We interpret these two features into the object oriented paradigm as the fact that social roles are objects, which are defined in and exist only inside other objects (corresponding to institutions), and that, through a role, external objects playing the role can access to the object (institution) the role belongs to.
This paper is about control mechanisms for virtual organizations. As a case study, we discuss the... more This paper is about control mechanisms for virtual organizations. As a case study, we discuss the Renewables Obligation (RO), a control mechanism that was introduced in the United Kingdom to stimulate the production of renewable energy. We apply a conceptual model based on normative multiagent systems (NMAS). We propose to model both the participants and the normative system as autonomous agents, having beliefs and goals. Norms, which can be internalized by the agents as obligations, are translated into conditional beliefs and goals of the normative system, which concern both detection and sanctioning measures. We show that the model can handle both the regulative and the evidential aspects of the case.
Proceedings of the fifth international joint conference on Autonomous agents and multiagent systems - AAMAS '06, 2006
In this paper we illustrate how a role-based semantics for agent communication languages can embe... more In this paper we illustrate how a role-based semantics for agent communication languages can embed the two main existing models of agent communication languages, respectively based on 'mental attitudes' and 'social commitments' semantics. These two models have been presented as incompatible approaches, but recently we illustrated for persuasion dialogues and using our normative multiagent systems framework, that they can be seen also as complimentary ones. Independently from our own multi-agent model, in this paper we illustrate for the speech act 'inform' how the role based semantics embeds the other two semantics.
There are two main traditions in defining a semantics for agent communication languages, based ei... more There are two main traditions in defining a semantics for agent communication languages, based either on mental attitudes or on social commitments. In this paper, we translate both traditions in a different approach in which the dialogue state is represented by the beliefs and goals publicly attributed to the roles played by the dialogue participants. On the one hand, this approach avoids the problems of mentalistic semantics, such as the unverifiability of private mental states. On the other hand, it allows to reuse the logics and implementations developed for FIPA compliant approaches.
We present an algorithm for replanning in a reactive agent architecture which incorporates decisi... more We present an algorithm for replanning in a reactive agent architecture which incorporates decision-theoretic notions to drive the planning and metadeliberation process. The deliberation component relies on a refinement planner which produces plans with optimal expected utility. The replanning algorithm we propose exploits the planner's ability to provide an approximate evaluation of partial plans: it starts from a fully refined plan and makes it more partial until it finds a more partial plan which subsumes more promising refinements; at that point, the planning process is restarted from the current partial plan.
Uploads
Papers by Guido Boella