Marco Colombetti

• Professor
• Professor (Full) at Politecnico di Milano

This paper addresses the problem of proposing a model of norms and a framework for automatically computing their violation or fulfilment. The proposed model can be used to express abstract norms able to regulate classes of actions that should or should not be performed in a temporal interval. We show how the model can be used to formalize obligatio...

This paper addresses the problem of proposing a model of norms and a framework for automatically computing their violation or fulfilment. The proposed T-NORM model can be used to express abstract norms able to regulate classes of actions that should or should not be performed in a temporal interval. We show how the model can be used to formalize ob...

https://people.lu.usi.ch/fornaran/AIComm2019/AIComm32(2019)_Fornara.pdf

Nowadays the studies on the formalization, enforcement, and monitoring of policies and norms is crucial in different fields of research and in numerous applications. ODRL 2.2 (Open Digital Right Language) is a W3C standard policy expression language formalized using semantic web technologies. It is used to represent permitted and prohibited actions...

Nowadays economy is every day more and more a digital economy where many human activities are performed by means of digital devices. Those digital activities produce and operate on a big amount of digital assets, as the data stored in datasets, documents, images, videos or audio files. Rationally, it is useless that digital assets are made public w...

In this paper we outline a theory of human deontology from a naturalistic perspective. In doing so we aim to explain how human beings deal with deontic relations (like obligations and rights) thanks to a specialised psychological infrastructure, which evolved to support human cooperation. This infrastructure includes a repertoire of emotions that p...

In this work, given the context of the driver, of the vehicle and of the environment, our objective is to correctly recognize the traffic situation and provide the driver with the corresponding assistance by providing notification or alert about the situation or the infraction that was committed, or acting directly on the vehicle. To do so, we need...

We present an approach for evaluating XACML policies using OWL technologies and DL reasoning. We explain how policies can be mapped to an OWL axiomatization, and how it is possible to generate answers to access requests using standard DL reasoning tools, all of that in the context of a complete XACML-compliant framework. Our model represents a subs...

Building on our previous works on human normativity, based on the concept of responsibility, we argue that: (i), the psychological infrastructure allowing people to enter relationships of responsibility is rooted in human affectivity, and in particular in certain types of emotions; and (ii), that different aspects of responsibility, and with them d...

We present a Description Logics approach to the management of XACML policies. We explain how policies can be mapped to a DL axiomatization, and how authorization requests can be answered using standard DL reasoning tools. Our model represents a valid substratum for managing policies whose expressivity can not be captured by standard engines. Furthe...

We present an XACML policy framework implementation using OWL and reasoning technologies. Reasoning allows to easily generate policy decisions in complex environments for expressive policies, while satisfying the requirements of reliability and consistency for the framework. Furthermore, OWL ontologies represent a valid substratum for tackling adva...

We compare a number of influential approaches to human communication with the aim of understanding what it means for interpersonal communication to be a form of social action. In particular, we discuss the large-scale social normativity advocated by speech act theory, the view of communication as small-scale social interaction proper of Gricean app...

When people interact in everyday situations, they constantly create new fragments of social reality: they do so when they make promises or agreements, but also when they submit requests or answer questions, when they greet each other or express gratitude. This type of social reality ‘in the small,’ that we call interpersonal reality, is deontic in...

We understand interpersonal reality as consisting of those social facts that are informally created by people for themselves in everyday interactions, and involve the collective acceptance of positive and negative deontic powers. We submit that, in the case of interpersonal reality, Gilbert’s concept of a joint commitment is a suitable view of what...

We propose an infrastructure for the design and development of Open Interaction Systems (OISs), based on solutions from Service Oriented Architecture, Semantic Technologies, and Normative Multiagent Systems. OISs are open to diverse types of participants (software agents), and enable them to interact with each other to achieve their objectives. To...

This paper integrates the responses to a set of questions from a distinguished set of panelists involved in a discussion at the Agreement Technologies workshop in Cyprus in December 2009. The panel was concerned with the relationship between the research areas of semantics, norms, and organizations, and the ways in which each may contribute to the...

Increasingly, software engineering involves open systems consisting of autonomous and heterogeneous participants or agents who carry out loosely coupled interactions. Accordingly, understanding and specifying communications among agents is a key concern. A focus on ways to formalize meaning distinguishes agent communication from traditional distrib...

According to Margaret Gilbert, a joint commitment (JC) is a commitment of two or more agents, called the parties of the JC, to engage in a common project. Creating a JC often involves an explicit agreement, carried out in a conversational interaction through overt communication. We explored aspects of such interactions that can be considered as com...

In Cognitive pragmatics: The mental processes of communication (2011), Bruno Bara presents a detailed summary of a theory of human communication, called “cognitive pragmatics,” which he has been developing since the 1980s together with a number of colleagues, and has been presented in several scientific articles and in a recent book (Bara 2010). Th...

The design and implementation of open interaction systems is widely recognized to be a crucial issue in the development of innovative applications on the Internet. In this paper we pursue the goal of enhancing interoperability and openness in open interaction systems by systematic use of web standards. We propose a way of using the semantic web lan...

In this short contribution we explain how our research has evolved from the publication of the following paper [2] with respect to its relevant aspects. This paper proposes a model of norms whose content is related to time, which are specified at design time and therefore are expressed in terms of roles played by the agents. Those norms have an act...

The design and implementation of open interaction systems (OIS) is widely recognized to be a crucial issue in the development of innovative applications on the Internet. In our past works we have pro-posed to specify the high-level component of an OIS by means of a set of artificial institutions. In this paper we describe the lower layers of a fram...

We propose a theory of communicative interactions based on the idea that it is constitutive of interpersonal communication to create and manage a fragment of social reality. We define such a fragment in terms of joint commitments of the interactants, and analyze how these commitments are made in a conversation. We distinguish between three layers o...

Monitoring the temporal evolution of obligations and prohibitions is a crucial aspect in the design of open interaction systems. In this paper we regard such obligations and prohibitions as cases of social commitment with starting points and deadlines, and propose to model them in OWL, the logical language recommended by the W3C for Semantic Web ap...

In this paper we want to reconcile two apparently conflicting intuitions: the first is that what a speaker means is just a function of his or her communicative intentions, independently of what the hearer understands, and even of the actual existence of a hearer; the second is that when communication is carried out successfully, the resulting meani...

The speciflcation and monitoring of conditional obligations and prohibitions with starting points and deadlines is a crucial aspect in the design of open interaction systems. In this paper we regard such obligations and prohibitions as cases of social commitment, and pro- pose to model them in OWL, the logical language recommended by the W3C for Se...

Agent Communication Languages (ACLs) have recently acquired a primary role in open multiagent systems, which need a standard communication framework shared by all interacting heterogeneous agents. According to the most important ACL standard proposals so far, agents are supposed to carry out the communication process by performing actions of a spec...

Learning objects paradigm is widely adopted in e-learning environments. Learning objects management can be improved using semantic technologies from ontology engineering and the semantic web. In this article we use a semantic model of the repository to improve both learning objects retrieval and composition. The use of domain knowledge enables auto...

The specification of open interaction systems is widely recognized to be a crucial issue, which involves the problem of finding a standard way of specifying: a communication language for the interacting agents, the entities that constitute the context of the interaction, and rules that regulate interactions. An approach to solve these problems cons...

Institutions have been proposed to explicitly represent norms in open multi-agent systems, where agents may not follow them and which therefore require mechanisms to detect violations. In doing so, they increase the efficiency of electronic transactions carried out by agents, but raise the problem of ensuring that such institutions are not characte...

We delineate a theory of communicative acts as situated ac-tions, through which agents co-construct a viable situation by creating or otherwise manipulating deontic affordances. We rely on Gilbert's theory of plural subjects to introduce the concept of joint meaning as a type of joint commitment. We then show that our approach allows for an innovat...

Institutions have been proposed to explicitly represent norms in open multi-agent systems, where agents may not follow them and which therefore require mechanisms to detect violations. In doing so, they increase the efficiency of electronic transactions carried out by agents, but raise the problem of ensuring that such institutions are not characte...

The specification of open interaction systems is widely recognized to be a crucial issue, which involves the problem of finding a standard way of specifying: a communication language for the interacting agents, the entities that constitute the context of the interaction, and rules that regulate interactions. An approach to solve these problems cons...

Robots that are capable of learning new tasks from humans need the ability to transform gathered abstract task knowledge into their own representation and dimensionality. New task knowledge that has been collected e.g. with Programming ...

In this paper we investigate two important and related aspects of the formalization of open interaction systems: how to specify norms, and how to enforce them by means of sanctions. The problem of specifying the sanctions associated with the violation of norms is crucial in an open system because, given that the compliance of autonomous agents to o...

Semantic annotations could improve the Legacy Web by ad- ding semantics to information which has already been pub- lished in form of unstructured text. Semi-automatic annota- tion tools seem the most viable way to obtain a contribution from users without requiring them to have a deep knowl- edge about semantics, however the effort to make them work...

Software agents’ ability to interact within different open systems, designed by different groups, presupposes an agreement on an unambiguous definition of a set of concepts, used to describe the context of the interaction and the communication language the agents can use. Agents’ interactions ought to allow for reliable expectations on the possible...

The Agent Communication Language (ACL) proposed by the Foundation for Intelligent Physical Agents (FIPA) is the most complete attempt to create a universally accepted standard so far. Nevertheless, this standard shows some shortcomings which are probably hindering an even greater impact upon the scientific research dealing with multiagent systems....

Since the beginning of the Nineteen-eighties, cognitive scientists have shown increasing interest in a range of phenomena, processes and capacities underlying human interaction, collectively referred to as intersubjectivity. The goal of this line of research is to give an account of the various forms of human interaction, and in particular of the a...

Norms deflned by institutions and enforced by organizations have been put forward as a mechanism to increase the e-- ciency and reliability of electronic transactions carried out by agents in open systems. Despite several approaches have been proposed to model protocols in terms of institutional concepts (e.g., obligations and powers) and to monito...

Institutions have been proposed as a means to regulate open interaction systems by introducing a set of norms (involving deontic positions like authorizations, obligations, prohibitions, and permissions) and to define the ontology of the context in which agents interact. To better clarify the interdependence existing among deontic positions and the...

In this paper we propose an application-independent model for the definition of artificial institutions that can be used to define open multi-agent sys-tems. Such a model of institutional reality makes us able also to define an objective and external semantics of a commitment-based Agent Communication Language (ACL). In particular we propose to reg...

We propose a new approach to integrate the navigation interface of a folksonomy adding explicit semantics provided by an ontology. We describe a tool that uses WordNet to build a semantic hierarchy of related tags that helps users find related resources in del.icio.us. In this way it is possible to combine the advantages of the traditional approach...

In this paper we enrich FIEVeL (a modelling language for institutions amenable to model checking) with new constructs to describe norms and sanc- tions. Moreover, we present a specication language to reason about the effective- ness of norms and sanctions in shaping agent interactions. Finally we show that when properties of articial institutions r...

As the volume of information in the read-write Web increases rapidly, folksonomies are becoming a widely used tool to organize and categorize resources in a bottom up, flat and inclusive way. However, due to their very structure, they show some drawbacks; in particular the lack of hierarchy bears some limitations in the possibilities of searching a...

Our research aims at providing an alternative model of agent communication to the one proposed by the Foundation for Intelligent Physical Agents (FIPA). We adopt the mainstream perspective that views agent communication as the performance of communicative acts, but we shift the focus from agents' mental states to their social state. Starting from t...

The notion of artificial institution is crucial for the specification of open and dynamic interaction frameworks where heterogeneous and autonomous agents can interact to face problems in various fields. In our view the spe cifi- cation of artificial institutions requires a clear standard definition of some ba sic concepts: the notion of ontology,...

This paper presents ANEMONE, a multi-agent platforms network that provides services for the academic community implemented by using the JADE agent development framework. In particular, ANEMONE provides a set of services to support i) academic people in some of their recurrent activities (fix an appointment, organize a meeting and search documents o...

The notion of artificial institution is crucial for the specification of open and dynamic interaction frameworks where heterogeneous and autonomous agents can interact to face problems in various fields, like for instance electronic commerce, business-to-business (B2B) applications, and personal assistant applications. In our view the specification...

In this paper we propose an operational method for the def- inition of the semantics of Agent Communication Languages based on the notion of social commitment. Our proposal is suitable for open inter- action frameworks where agents, designed by independent constructors, dynamically enter and leave dierent interaction systems. In this type of enviro...

The ability to express temporal conditions, like for example deadlines, is extremely important in agent applications. Nevertheless, communication standards like FIPA ACL do not outline a uniform way to specify such conditions in Content Language expressions. In this paper we extend a CTL*-like temporal language with two very expressive interval ope...

The current ACL proposals show some shortcomings with respect to the definition of their semantics. Our paper aims at tackling those issues by defining an ACL semantics as a specification of the analytical effects of agent communicative acts. We analyze agent communication in terms of concepts taken from Speech Act Theory, as several researchers ha...

In this paper we propose a model of agents' institutional reality that makes us able to define the semantics of a set of communicative acts suitable for agent communication. Taking inspiration from Searle's studies on the construction of social reality, we introduce our notion of Artificial Institution consisting of: an ontology of the concepts def...

Nelle scienze cognitive prevale da un quarto di secolo il paradigma computazionale: si assume cioè che i processi mentali possano essere descritti e spiegati in quanto processi di computazione. Questo punto di vista non è privo di aspetti problematici, che diventano particolarmente evidenti quando l’oggetto di studio è la coscienza: la stessa locuz...

In this paper we propose to regard an Agent Communication Language (ACL) as a set of conventions to act on a fragment of institutional reality, defined in the context of an artificial institution. Within such an approach, we first reformulate a previously proposed commitment-based semantics for ACLs. In particular we show that all commonly used typ...

This paper aims at defining the semantics of Agent Communication Languages (ACLs) in terms of changes in the social relationships between agents, represented in terms of social commitments. We take commitment to be a primitive concept underlying the social dimension of multiagent systems, and define a basic artificial institution that provides agen...

We propose a method for the definition of interaction protocols to be used in open multiagent systems. Starting from the assumption that language is the fundamental component of every interaction, we first propose a semantics for Agent Communication Languages based on the notion of social commitment, and then use it to define the meaning of a set o...

As part of the goal of developing a genuinely open multiagent system, many efforts are devoted to the definition of a standard Agent Communication Language (ACL). The aim of this paper is to propose a logical framework for the definition of ACL semantics based upon the concept of (social) commitment. Our framework relies on the assumption that agen...

We propose a method for the definition of interaction protocols to be used in open multiagent systems. Starting from the assumption that language is the fundamental component of every interaction, we first propose a semantics for Agent Communication Languages based on the notion of social commitment, and then use it to define the meaning of a set o...

We propose a method for the definition of interaction protocols to be used in open multiagent systems. Starting from the assumption that language is the fundamental component of every interaction, we first propose a semantics for Agent Communication Languages based on the notion of social commitment, and then use it to define the meaning of a set o...

The development of open multiagent systems presupposes not only a standard Agent Communication Language and common conversation protocols, but also some kind of social framework within which agent interactions can be carried out.

As supply chain networks are becoming more and more global, process coordination must be considered a crucial point for successful business management. The need for a suitable management and communication framework is thus becoming evident. We already have some examples showing that information sharing is a key-point at certain levels of a supply c...

In this paper we propose an operational method to express the meaning of the messages exchanged among agents that interact in open environments. In an open environment, like for example the Internet, agents are usually designed by different constructors, so it is very important to define the meaning of a standard, widely accepted Agent Communicatio...

All major proposals in the field of Agent Communication Languages deal with agent communication in terms of speech acts. This choice is important not only because it allows one to rely on a powerful and deep theory of communication, namely Speech Act Theory, but also because AI has developed computationally effective ways of dealing with actions. H...

All major proposals in the field of Agent Communication Languages deal with agent communication in terms of speech acts. This choice is important not only because it allows one to rely on a powerful and deep theory of communication, namely Speech Act Theory, but also because AI has developed computationally effective ways of dealing with actions. H...

Software agents are starting to play a significant role in the field of electronic commerce. Particularly, as mediators they must be completely trusted by the user. This paper tackles this subject by proposing a practically tailored formal model, based on BDI. This model is tested on the ground, firstly, by describing an existing agent for electron...

An agent is a system that interacts with an environment continually and without human assistance in order to carry out a predefined task. We are interested in developing artificial agents that act rationally, in the sense that they are able to maximize a suitable utility function. In this chapter, we describe the main problems underlying the realiz...

Most approaches to agent communication are based on some notion of communicative act. I argue that a satisfactory model of communication should deal with three related but conceptually independent aspects of communicative acts: the semantic aspect, concerning the literal meaning of communicative acts; the normative aspect, concerning the obligation...

This paper presents the main elements of Albatross, an agent communication language whose definition is currently under development. The semantics of Albatross, based on the social notion of commitment, allows one to define speech act types in a neat and concise way. I describe the logical relationship between messages and speech acts; give sample...

. In the last few years we have used ALECSYS, a parallel learning classifier system based on the genetic algorithm, to develop behavioral modules for mobile robots, both simulated and real. In this paper we briefly report on our experience, and then reflect on various concepts stemming from the application of evolutionary computation to agent build...

In this paper we propose a three-stage incremental approach to the development of autonomous agents. We discuss some issues about the characteristics which differentiate reinforcement programs (RPs), and define the trainer as a particular kind of RP. We present a set of results obtained running experiments with a trainer which provides guidance to...

Industrial automation calls for behavioral intelligence, that is, a mixture of flexibility, robustness and adaptiveness of robot behavior. We argue that efficient machine learning techniques can be a valuable tool for achieving behavioral intelligence. As a case study we apply ALECSYS, an implementation of a learning classifier system on a net of t...

In this paper we present an application of ALECSYS, a distributed learning classifier system, to the control of a robot arm. ALECSYS is initialised with a set of randomly generated rules and is trained to control a robot arm whose task is to reach a non moving light source. At this point of our research our results are relative to the simulation of...

I propose a propositional modal logic of intentional communication, a circular concept which develops an idea by Airenti, Bara and Colombetti (1993). A communication operator is added to a multi-modal language of individual belief, common belief and intention, and its possible world semantics is justified through a fixpoint construction. A normal m...

We asked ‘What is a Learning Classifier System’ to some of the best-known researchers in the field. These are their answers.

We analyze XCS learning capabilities in stochastic environments where the result of agent actions can be uncertain. We show that XCS can cope when the degree of uncertainty is limited. We propose an extension to XCS, called XCSm, which can learn optimal solutions for higher degrees of uncertainty. We test XCSm when the uncertainty affects the whole...

We propose a modular architecture for autonomous robots which allows for the implementation of basic behavioral modules by both programming and training, and accommodates for an evolutionary development of the interconnections among modules. This architecture can implement highly complex controllers and allows for incremental shaping of the robot b...

Agent technology is rapidly gaining favor among computer scientists. In typical applications, agents are not viewed as isolated systems, but rather as individuals living in a complex society of other agents and interacting with them, in order to cooperate to the achievement of tasks and negotiate solutions to conflictual situations. Social agents o...

We present some results of our research in the field of Machine Learning applied to robotics problems. In particular we have investigated on: (i) the application of Learning Classifier Systems to the synthesis of robot controllers; (ii) learning of fuzzy controllers; (iii) learning of purposeful representations of the environment; (iv) and the appl...

This paper is concerned with training an agent to perform sequential behavior. In previous work we have been applying reinforcement learning techniques to control a reactive robot. Obviously, a pure reactive system is limited in the kind of interactions it can learn. In particular, it can only learn what we call pseudo-sequences, that is sequences...

Learning plays a vital role in the development of situated agents. In this paper, we explore the use of reinforcement learning to "shape" a robot to perform a predefined target behavior. We connect both simulated and real robots to ALECSYS, a parallel implementation of a learning classifier system with an extended genetic algorithm. After classifyi...

The mechanisms of learning, and the complex balance between what is learned and what is genetically determined, are one of the main concerns of behavioral sciences. It is our opinion that a similar situation arises in autonomous robotics, where a major problem is in deciding what should be explicitly designed and what should be left for the robot t...

foreword by Lashon Booker To program an autonomous robot to act reliably in a dynamic environment is a complex task. The dynamics of the environment are unpredictable, and the robots' sensors provide noisy input. A learning autonomous robot, one that can acquire knowledge through interaction with its environment and then adapt its behavior, greatly...

This paper describes QLB, a quantified logic of belief that is a possible extension of the modal system KD45n to predicate level. The main features of QLB are that: (i) it is allowed to quantify over the agents of belief; (ii) the belief operator can be indexed by any term of the formal language; (iii) terms are not rigid designators, but are inter...

ociator which can only learn linearly separable functions, while ALECSYS is a more general system which does not have this limit: a possible lower speed of ALECSYS is counterbalanced by its greater generality. Second, ALECSYS learns by reinforcement learning, while Nehmzow and McGonigle's pattern associator learns by supervised learning. Again, ALE...

Robot Shaping: An Experiment in Behavior Engineering, by Marco Dorigo and Marco Colombetti. Cambridge, MA: MIT Press/Bradford Books, forthcoming October 1997. Approximately 200 pages, including 87 figures and 15 tables, approximately $40. ISBN 0-262-04164-2.

We propose Behavior Engineering as a new technological area whose aim is to provide methodologies and tools for developing autonomous robots. Building robots is a very complex engineering enterprise that requires the exact definition and scheduling of the activities which a designer, or a team of designers, should follow. Behavior Engineering is, w...

In this paper we present an example of the application of a technique, which we call robot shaping, to designing and building learning autonomous robots. Our autonomous robot (called HAMSTER 1 ) is a multi-sensor mobile robot that performs the task of collecting "food" and bringing it to its "nest". Its control architecture is based on the behavior...

Reinforcement learning (RL) is a direct control method and does not require an explicit model of the system to be controlled. In RL, the control system is developed by using as a unique source of information a scalar, the so-called reinforcement, which evaluates control actions. The control system receives either positive or negative reinforcements...

Learning plays a vital role in the development of autonomous agents. In this paper, we explore the use of reinforcement learning to “shape” a robot to perform a predefined target behavior. We connect both simulated and real robots to Alecsys, a parallel implementation of a learning classifier system with an extended genetic algorithm. After classif...