Theory of mind refers to the attribution of mental states that humans ascribe to other humans or objects (such as computer-based systems). Recently, the attribution of mental states has been investigated toward Artificial Intelligence (AI) as a basic manner to capture people's engagement toward it, and people's perception about AI social skills and AI capabilities. In line with this idea, mental state attribution can be used as an indirect measure of students' understanding of AI functioning, and in particular of the kind of interactions students may have with AI systems. Too often is the case of people using generative AI systems in ways that exceed their actual ways of functioning. In our study, children of age in the range 9-12 were involved in one-shot unplugged activities concerning data and models. The unplugged activities were not aimed at teaching the theory of Machine Learning, but rather they were designed so as to provide awareness on some basic mechanisms and help developing a correct use of tools that are becoming more and more present in everyday life. This paper introduces the activities and reports the results that were achieved.

Accountability is a central notion in both political science and software engineering, where it underpins trust, transparency, and responsibility in complex systems. Multi-agent systems (MAS) provide fertile ground for investigating accountability as a first-class computational construct. While earlier approaches in MAS research relegated accountability to an implicit property, recent extensions model it explicitly. This paper explores accountability in the well-known multi-agent JaCaMo framework as an organizational, normative, and structural mechanism. We argue that formalizing accountability in MAS provides a robust foundation for legal reasoning, compliance verification, and transparent governance of complex distributed systems, ensuring traceable responsibilities and structured feedback flows. We introduce its conceptual foundations, the JaCaMo implementation, and its implications for legal knowledge systems, showing how accountability provides a formal bridge between normative theories and executable systems.

Artificial Intelligence is undoubtedly becoming pervasive in everyday life of everyone. In this setting, developing correct AI conception since childhood is not only a need to be addressed in educational curricula, but is also a children right. Accordingly, several initiatives at national and international levels aim at promoting AI and emerging technology literacy, supported also by a proliferation in the literature of learning courses covering a variety of topics, learning objectives and targeted ages. Schools are therefore pushed to introduce innovative activities for children in their curricula. In this paper, we report the results of a case study where we tested the contribution of an AI block-based course in developing computational thinking, and human and AI minds understanding in fifth and sixth grade children.

Depopulation and the consequent erosion of local expertise hinder the sustainable growth of the mountain areas. The reduction of work force and of talent negatively affects the local enterprises, and in particular their power of innovation, creating a divide with those that are located in the urban centers, whose greater richness of services and of choice creates a more fertile, and thus more attractive, environment especially for young people. In order to invert the trend, there is the need to create the conditions for triggering social innovation, that is, to let new strategies, initiatives, processes, or organizations emerge that meet social needs and that cause radical changes in the social system. The proposal presented in this paper embraces this assumption and explores a novel approach, consisting in the establishment of specialized networks of co-working spaces as a regional development tool. Specifically: 1) we describe a socio-technical framework supporting the creation of networks of co-working spaces and we model its effects by way of game theory; 2) we elicit a set of requirements that can guide the development of an actual platform, supporting the creation of networks of co-working spaces, which can flexibly accommodate many organizational models.

Process alignment aims at establishing a matching between a process model run and a log trace. To improve such a matching, process alignment techniques often exploit contextual conditions to enable computations that are more informed than the simple edit distance between model runs and log traces. The paper introduces a novel approach to process alignment which relies on contextual information expressed as responsibilities . The notion of responsibility is fundamental in business and organization models, but it is often overlooked. We show the computation of optimal alignments can take advantage of responsibilities. We leverage on them in two ways. First, responsibilities may sometimes justify deviations. In these cases, we consider them as correct behaviors rather than errors. Second, responsibilities can either be met or neglected in the execution of a trace. Thus, we prefer alignments where neglected responsibilities are minimized. The paper proposes a formal framework for responsibilities in a process model, including the definition of cost functions for computing optimal alignments. We also propose a branch-and-bound algorithm for optimal alignment computation and exemplify its usage by way of two event logs from real executions.

Ethics by Design concerns the methods, algorithms and tools needed to endow autonomous agents with the capability to reason about the ethical aspects of their decisions, and the methods, tools and formalisms to guarantee that an agent's behavior remains within given moral bounds. In this context some questions arise: How and to what extent can agents understand the social reality in which they operate, and the other intelligences (AI, animals and humans) with which they co-exist? What are the ethical concerns in the emerging new forms of society, and how do we ensure the human dimension is upheld in interactions and decisions by autonomous agents?. But overall, the central question is: "Can we, and should we, build ethically-aware agents?" This paper presents initial conclusions from the thematic day of the same name held at PRIMA2017, on October 2017.

We present the JaCaMo+ framework for programming multiagent systems (MAS), where agents interact thanks to commitment-based interaction protocols. Commitment protocols are realized as artifacts that maintain a social state and notify to the participating agents those events that are relevant to the interaction. We discuss the advantages, like increased modularity and flexibility, that are brought by commitment-ruled interactions with respect to other proposals. We trace back such advantages to the possibility of relying on a standardized commitment lifecycle. We explain how to use the framework to program interacting agents by using the Netbill protocol as running example, and the Gold Miners scenario as a more complex programming example.

Computational Ethics and Accountability are becoming topics of increasing societal impact; in particular, on the one hand, in the context of recent advances in AI and machine-learning techniques, people and organizations accept decisions made for them by machines, be they buy-sell decisions, pre-filtering of applications, deciding which content users are presented, which personal data are shared and used by third parties, up to automated driving.In each of these application scenarios, where algorithms and machines support or even replace human decisions, ethical issues may arise.On the other hand, algorithms and machines can play the role of verifying and cross-checking compliance of human players in analyzing digital records of social interactions, for instance, in business transactions and processes, but also in following rules of conduct in online social interactions.Closing the circle, based on such checks, again automated decisions may be implied that involve ethical requirements, such as nondiscrimination and fairness.Apart from infamous "trolley problems" (Edmonds 2013), where even philosophers struggle to judge what is the "right" decision and going either way has dramatic impacts, there are more subtle everyday decisions that we now either happily delegate to machines or that are digitally recorded, which may have ethical implications: handling of personal data has to follow strict legal regulations, especially in social networks and in re-sharing personal data with businesses, (social) norms should be followed also in domains where automated agents enter interactions that were typically executed by human actors only, and fair business practices should be ensured within business processes, compliant with regulations, laws, and best practices.In all these areas, at the very least, we expect transparency and accountability from automated decision and decision support systems: that is, we should require these systems to be transparent about how they make decisions and knowing who is accountable for those decisions and their effects.Many voices demand a more responsible technology and engineering approach, such as articulated in the Copenhagen Letter (Techfestival 2017), or recent initiatives to standardize value-based ethically compliant system design, such as IEEE's P7000 (for Ethical Life-Cycle Concerns Working Group (EMELC-WG) 2017) family of standards.

Social Computing (SC) requires agents to reason seamlessly both on their social relationships and on their goals, beliefs. We claim the need to explicitly represent the social state and social relationships as resources, available to agents. We built a framework, based on JaCaMo, where this vision is realized and SC is implemented through social commitments and commitment protocols.

Socio-Technical Systems demand an evolution of computing into social computing, with a transition from an individualistic to a societal view. As such, they seem particularly suitable to realize multiparty, cross-organizational systems. Multi-Agent Systems are a natural candidate to realize Socio-Technical Systems. However, while Socio-Technical Systems envisage an explicit layer that contains the regulations that all parties must respect in their interaction, and thus preserve the agents' autonomy, current frameworks and platforms require to hard-code the coordination requirements inside the agents. We propose to explicitly represent the missing layer of Socio-Technical Systems in terms of social relationships among the involved parties, that is, in terms of a set of normatively defined relationships among two or more parties, subject to social control by monitoring the observable behaviour. In our proposal, social relationships are resources, available to agents, who use them in their practical reasoning. Both agents and social relationships are first-class entities of the model. The work also describes 2COMM4JADE, a framework that realizes the proposal by extending the well-known JADE and CArtAgO. The impact of the approach on programming is explained both conceptually and with the help of an example.

Interaction protocols play a fundamental role in multiagent systems. In this work, after analyzing the trends that are emerging not only from research on multiagent interaction protocols but also from neighboring fields, like research on workflows and business processes, we propose a novel definition of commitment-based interaction protocols, that is characterized by the decoupling of the constitutive and the regulative specifications and that explicitly foresees a representation of the latter based on constraints among commitments. A clear distinction between the two representations has many advantages, mainly residing in a greater openness of multiagent systems, and an easier reuse of protocols and of action definitions. A language, named 2CL, for writing regulative specifications is also given together with a designer-oriented graphical notation.

The proposal of Elisa Marengo's thesis is to extend commitment protocols in order to (i) allow for expressing commitments to temporal regulations, and (ii) to supply a tool for expressing laws, conventions and the like, in order to specify legal interactions.These two aspects will be deeply investigated in the proposal of a unified framework.This proposal is part of ongoing work that will be included in the thesis.

This paper presents a system which learns and maintains a body of heuristic rules, useful to decide about the behaviour of a robot, from apriori knowledge and a set of examples. The apriori knowledge consists of a causal model of the domain, stating the relation- ships among basic phenomena, and a body of phe- nomenological theory, describing the links between ab- stract concepts and their possible manifestations in the world. The phenomenological knowledge is used deduc- tively, the causal model is used abductively and the examples are used inductively. The problems of imper- fection and intractability of the theory are handled by allowing the system to make assumptions during its reasoning. In this way, robust knowledge can be learned with limited complexity and limited number of exam- ples. For the sake of illustration, an simple example where a robot is asked to learn whether an object is dangerous because hot, or not, is presented .