Mattia Fumagalli

AI technologies are ubiquitous, manifesting in various forms and applications. Nowadays, many of them are presented as being general purpose, though this sometimes translates into having unclear purposes. Within this landscape, establishing normative criteria for the functioning of (some of) these artefacts has grown increasingly complex. In this work, we advocate the use of teleological explanation as a suitable ally in addressing the challenges arising in this scenario. We defend the idea that teleological explanation is particularly useful in the case of artefacts in general, and we discuss how, with some adaptation, it can be leveraged to support the assessment of AI artefacts according to their purpose(s). Our contribution is thus threefold: (1) exploiting some of the assumptions adopted in teleological explanation to support the clarification of (general-purpose) AI artefacts’ purpose(s); (2) assisting in the comparison and assessment of AIs via metrics inspired by the literature concerning teleological explanation; and (3) providing insights for defining a unified framework for designing AI benchmarks.

Formal Ontology is a discipline whose business is to develop formal theories about general aspects of reality such as identity, dependence, parthood, truth-making, causality, etc. A foundational ontology is a specific consistent set of these ontological theories that support activities such as domain analysis, conceptual clarification, and meaning negotiation. A (well-founded) core ontology specifies, under a foundational ontology, the central concepts and relations of a given domain. Foundational and core ontologies can be seen as ontology engineering frameworks to systematically address the laborious task of building large (more specific) domain ontologies. However, both in research and industry, it is common that ontologies as computational artifacts are built without the aid of any framework of this kind, often yielding modeling mistakes and representation gaps. In this paper, we analyze a case in the domain of cybersecurity, namely, the case of D3FEND — an OWL knowledge graph of cybersecurity countermeasure techniques proposed by the MITRE Corporation. Based on the Reference Ontology for Security Engineering (ROSE), a core ontology of the security domain founded in the Unified Founda-tional Ontology (UFO), our investigation reveals a number of semantic issues and opportunities for improvement in D3FEND, including missing concepts, semantic overload of terms, and lacking constraints that cause an under-specification of the model. As a result of our ontological analysis, we propose several suggestions for the appropriate redesign of D3FEND to overcome those issues.

The modelling of concepts, besides involving disciplines like philosophy of mind and psychology, is a fundamental and lively research problem in several artificial intelligence (AI) areas, such as knowledge representation, machine learning, and natural language processing. In this scenario, the most prominent proposed solutions adopt different (often incompatible) assumptions about the nature of such a notion. Each of these solutions has been developed to capture some specific features of concepts and support some specific (artificial) cognitive operations. This paper critically reviews the most notable computational approaches to the representation of concepts. The main goals are (i) to provide a shared terminology for the desiderata of concepts and their computational representation; (ii) to classify and assess the heterogeneous computational approaches according to the provided terminology; (iii) to provide a reader who may not be very familiar with theories of concepts with an introduction to major themes in this research and with pointers to different research projects, and (iv) to offer philosophers, and potentially AI practitioners, a well-informed guide for selecting among various (and possibly competing) computational representations of concepts.

Risk propagation encompasses a plethora of techniques for analyzing how risk “spreads” in a given system. Albeit commonly used in technical literature, the very notion of risk propagation turns out to be a conceptually imprecise and overloaded one. This might also explain the multitude of modeling solutions that have been proposed in the lit- erature. Having a clear understanding of what exactly risk is, how it be quantified, and in what sense it can be propagated is fundamental for devising high-quality risk assessment and decision-making solutions. In this paper, we exploit a previous well-established work about the nature of risk and related notions with the goal of providing a proper interpre- tation of the different notions of risk propagation, as well as revealing and harmonizing the alternative semantics for the links used in common risk propagation graphs. Finally, we discuss how these results can be leveraged in practice to model risk propagation scenarios.

In Risk Management, security issues arise from complex relations among objects and agents, their capabilities and vulnerabilities, the events they are involved in, and the value and risk they ensue to the stakeholders at hand. Further, there are patterns involving these relations that crosscut many domains, ranging from information security to public safety. Understanding and forming a shared conceptualization and vocabulary about these notions and their relations is fundamental for modeling the corresponding scenarios, so that proper security countermeasures can be devised. Ontologies are instruments developed to address these conceptual clarification and terminological systematization issues. Over the years, several ontologies have been proposed in Risk Management and Security Engineering. However, as shown in recent literature, they fall short in many respects, including generality and expressivity - the latter impacting on their interoperability with related models. We propose a Reference Ontology for Security Engineering (ROSE) from a Risk Treatment perspective. Our proposal leverages on two existing Reference Ontologies: the Common Ontology of Value and Risk and a Reference Ontology of Prevention, both of which are grounded on the Unified Foundational Ontology (UFO). ROSE is employed for modeling and analysing some cases, in particular providing clarification to the semantically overloaded notion of Security Mechanism.

In recent years, the design and production of information systems have seen significant growth. However, these information artefacts often exhibit characteristics that compromise their reliability. This issue appears to stem from the neglect or underestimation of certain crucial aspects in the application of Information Systems Design (ISD). For example, it is frequently difficult to prove when one of these products does not work properly or works incorrectly (falsifiability), their usage is often left to subjective experience and somewhat arbitrary choices (anecdotes), and their functions are often obscure for users as well as designers (explainability). In this paper, we propose an approach that can be used to support the analysis and re-(design) of information systems grounded on a well-known theory of information, namely, teleosemantics. This approach emphasizes the importance of grounding the design and validation process on dependencies between four core components: the producer (or designer), the produced (or used) information system, the consumer (or user), and the design (or use) purpose. We analyze the ambiguities and problems of considering these components separately. We then present some possible ways in which they can be combined through the teleological approach. Also, we debate guidelines to prevent ISD from failing to address critical issues. Finally, we discuss perspectives on applications over real existing information technologies and some implications for explainable AI and ISD.