Background. Endowing intelligent systems with semantic data commonly requires designing and instantiating ontologies with domain-specific knowledge. Especially in the early phases, those activities are typically performed manually by human experts possibly leveraging on their own experience. The resulting process is therefore time-consuming, error-prone, and often biased by the personal background of the ontology designer. Objective. To mitigate that issue, we propose a novel domain-independent approach to automatically instantiate ontologies with domain-specific knowledge, by leveraging on large language models (LLMs) as oracles. Method. Starting from (i) an initial schema composed by inter-related classes and properties and (ii) a set of query templates, our method queries the LLM multiple times, and generates instances for both classes and properties from its replies. Thus, the ontology is automatically filled with domain-specific knowledge, compliant to the initial schema. As a result, the ontology is quickly and automatically enriched with manifold instances, which experts may consider to keep, adjust, discard, or complement according to their own needs and expertise. Contribution. We formalise our method in general way and instantiate it over various LLMs, as well as on a concrete case study. We report experiments rooted in the nutritional domain where an ontology of food meals and their ingredients is automatically instantiated from scratch, starting from a categorisation of meals and their relationships. There, we analyse the quality of the generated ontologies and compare ontologies attained by exploiting different LLMs. Experimentally, our approach achieves a quality metric that is up to five times higher than the state-of-the-art, while reducing erroneous entities and relations by up to ten times. Finally, we provide a SWOT analysis of the proposed method.

Abstract The expressive power and effectiveness of large language models (LLMs) is going to increasingly push intelligent agents towards sub-symbolic models for natural language processing (NLP) tasks in human–agent interaction. However, LLMs are characterised by a performance vs. transparency trade-off that hinders their applicability to such sensitive scenarios. This is the main reason behind many approaches focusing on local post-hoc explanations, recently proposed by the XAI community in the NLP realm. However, to the best of our knowledge, a thorough comparison among available explainability techniques is currently missing, as well as approaches for constructing global post-hoc explanations leveraging the local information. This is why we propose a novel framework for comparing state-of-the-art local post-hoc explanation mechanisms and for extracting logic programs surrogating LLMs. Our experiments—over a wide variety of text classification tasks—show how most local post-hoc explainers are loosely correlated, highlighting substantial discrepancies in their results. By relying on the proposed novel framework, we also show how it is possible to extract faithful and efficient global explanations for the original LLM over multiple tasks, enabling explainable and resource-friendly AI techniques.

Federated Learning (FL) represents the de-facto standard paradigm for enabling distributed learning over multiple clients in real-world scenarios. Despite the great strides reached in terms of accuracy and privacy awareness, the real adoption of FL in real-world scenarios, in particular in industrial deployment environments, is still an open thread. This is mainly due to privacy constraints and to the additional complexity stemming from the set of hyperparameters to tune when employing AI techniques on bandwidth-, computing-, and energy-constrained nodes. Motivated by these issues, we focus on scenarios where participating clients are characterised by highly heterogeneous computing capabilities and energy budgets proposing EneA-FL, an innovative scheme for serverless smart energy management. This novel approach dynamically adapts to optimize the training process while fostering seamless interaction between Internet of Things (IoT) devices and edge nodes. In particular, the proposed middleware provides a containerised software module that efficiently manages the interaction of each worker node with the central aggregator. By monitoring local energy budget, computational capabilities, and target accuracy, EneA-FL intelligently takes informed decisions about the inclusion of specific nodes in the subsequent training rounds, effectively balancing the tripartite trade-off between energy consumption, training time, and final accuracy. Finally, in a series of extensive experiments across diverse scenarios, our solution demonstrates impressive results, achieving between 30% and 60% lower energy consumption against popular client selection approaches available in the literature while being up to 3.5 times more efficient than standard FL solutions.

In the context of Information-Centric Networking, Interest Flooding Attacks (IFAs) represent a new and dangerous sort of distributed denial of service. Since existing proposals targeting IFAs mainly focus on local information, in this paper we propose GNN4IFA as the first mechanism exploiting complex non-local knowledge for IFA detection by leveraging Graph Neural Networks (GNNs) handling the overall network topology.In order to test GNN4IFA, we collect SPOTIFAI, a novel dataset filling the current lack of available IFA datasets by covering a variety of IFA setups, including ~40 heterogeneous scenarios over three network topologies. We show that GNN4IFA performs well on all tested topologies and setups, reaching over 99% detection rate along with a negligible false positive rate and small computational costs. Overall, GNN4IFA overcomes state-of-the-art detection mechanisms both in terms of raw detection and flexibility, and – unlike all previous solutions in the literature – also enables the transfer of its detection on network topologies different from the one used in its design phase.

Neuro-symbolic integration of symbolic and subsymbolic techniques represents a fast-growing AI trend aimed at mitigating the issues of neural networks in terms of decision processes, reasoning, and interpretability.Several state-of-theart neuro-symbolic approaches aim at improving performance, most of them focusing on proving their effectiveness in terms of raw predictive performance and/or reasoning capabilities.Meanwhile, few efforts have been devoted to increasing model trustworthiness, interpretability, and efficiency-mostly due to the complexity of measuring effectively improvements in terms of trustworthiness and interpretability.This is why here we analyse and discuss the need for ad-hoc trustworthiness metrics for neurosymbolic techniques.We focus on two popular paradigms mixing subsymbolic computation and symbolic knowledge, namely: (i) symbolic knowledge extraction (SKE), aimed at mapping subsymbolic models into human-interpretable knowledge bases; and (ii) symbolic knowledge injection (SKI), aimed at forcing subsymbolic models to adhere to a given symbolic knowledge.We first emphasise the need for assessing neuro-symbolic approaches from a trustworthiness perspective, highlighting the research challenges linked with this evaluation and the need for ad-hoc trust definitions.Then we summarise recent developments in SKE and SKI metrics focusing specifically on several trustworthiness pillars such as interpretability, efficiency, and robustness of neuro-symbolic methods.Finally, we highlight open research opportunities towards reliable and flexible trustworthiness metrics for neuro-symbolic integration.

Abstract Bridging intelligent symbolic agents and sub-symbolic predictors is a long-standing research goal in AI. Among the recent integration efforts, symbolic knowledge injection (SKI) proposes algorithms aimed at steering sub-symbolic predictors’ learning towards compliance w.r.t. pre-existing symbolic knowledge bases. However, state-of-the-art contributions about SKI mostly tackle injection from a foundational perspective, often focussing solely on improving the predictive performance of the sub-symbolic predictors undergoing injection. Technical contributions, in turn, are tailored on individual methods/experiments and therefore poorly interoperable with agent technologies as well as among each others. Intelligent agents may exploit SKI to serve many purposes other than predictive performance alone—provided that, of course, adequate technological support exists: for instance, SKI may allow agents to tune computational, energetic, or data requirements of sub-symbolic predictors. Given that different algorithms may exist to serve all those many purposes, some criteria for algorithm selection as well as a suitable technology should be available to let agents dynamically select and exploit the most suitable algorithm for the problem at hand. Along this line, in this work we design a set of quality-of-service (QoS) metrics for SKI, and a general-purpose software API to enable their application to various SKI algorithms—namely, platform for symbolic knowledge injection (PSyKI). We provide an abstract formulation of four QoS metrics for SKI, and describe the design of PSyKI according to a software engineering perspective. Then we discuss how our QoS metrics are supported by PSyKI. Finally, we demonstrate the effectiveness of both our QoS metrics and PSyKI via a number of experiments, where SKI is both applied and assessed via our proposed API. Our empirical analysis demonstrates both the soundness of our proposed metrics and the versatility of PSyKI as the first software tool supporting the application, interchange, and numerical assessment of SKI techniques. To the best of our knowledge, our proposals represent the first attempt to introduce QoS metrics for SKI, and the software tools enabling their practical exploitation for both human and computational agents. In particular, our contributions could be exploited to automate and/or compare the manifold SKI algorithms from the state of the art. Hence moving a concrete step forward the engineering of efficient, robust, and trustworthy software applications that integrate symbolic agents and sub-symbolic predictors.

Abstract We propose a novel method to inject symbolic knowledge in form of Datalog formulæ into neural networks (NN), called Knowledge Injection via Network Structuring (KINS). The idea behind our method is to extend NN internal structure with ad-hoc layers built out of the injected symbolic knowledge. KINS does not constrain NN to any specific architecture, neither requires logic formulæ to be ground. Moreover, it is robust w.r.t. both lack of data and imperfect/incomplete knowledge. Experiments are reported, involving multiple datasets and predictor types, to demonstrate how KINS can significantly improve the predictive performance of the neural networks it is applied to.

The success of neural networks (NNs) is tightly linked with their architectural design—a complex problem by itself. We here introduce a novel framework leveraging Graph Neural Networks to Generate Neural Networks (GNN2GNN) where powerful NN architectures can be learned out of a set of available architecture-performance pairs. GNN2GNN relies on a three-way adversarial training of GNN, to optimise a generator model capable of producing predictions about powerful NN architectures. Unlike Neural Architecture Search (NAS) techniques proposing efficient searching algorithms over a set of NN architec- tures, GNN2GNN relies on learning NN architectural design criteria. GNN2GNN learns to propose NN architectures in a single step – i.e., training of the generator –, overcoming the recursive approach characterising NAS. Therefore, GNN2GNN avoids the expensive and inflexible search of efficient structures typical of NAS approaches. Extensive experiments over two state-of-the-art datasets prove the strength of our framework, showing that it can generate powerful architectures with high probability. Moreover, GNN2GNN outperforms possible counterparts for generating NN architectures, and shows flexibility against dataset quality degradation. Finally, GNN2GNN paves the way towards generalisation between datasets.

Abstract In this paper we present the computational model of Arg2P, a logic-based argumentation framework for defeasible reasoning and agent conversation particularly suitable for explaining agent intelligent behaviours. The model is reified as the Arg2P technology, which is presented and discussed both from an architectural and a technological perspective so as to point out its potential in the engineering of intelligent systems. Finally, an illustrative application scenario is discussed in the domain of computable law for autonomous vehicles.

Abstract Precisely when the success of artificial intelligence (AI) sub-symbolic techniques makes them be identified with the whole AI by many non-computer-scientists and non-technical media, symbolic approaches are getting more and more attention as those that could make AI amenable to human understanding. Given the recurring cycles in the AI history, we expect that a revamp of technologies often tagged as “classical AI”—in particular, logic-based ones—will take place in the next few years. On the other hand, agents and multi-agent systems (MAS) have been at the core of the design of intelligent systems since their very beginning, and their long-term connection with logic-based technologies , which characterised their early days, might open new ways to engineer explainable intelligent systems . This is why understanding the current status of logic-based technologies for MAS is nowadays of paramount importance. Accordingly, this paper aims at providing a comprehensive view of those technologies by making them the subject of a systematic literature review (SLR). The resulting technologies are discussed and evaluated from two different perspectives: the MAS and the logic-based ones.

Edge Computing is rapidly gaining traction in scenarios such as Cyber-Physical Systems and Web of Things. Whereas the Cloud hides heterogeneity of devices behind its standard interfaces and protocols, the Edge should deal with it, as well as with embracing openness and governing interactions. In this paper we propose TuSoW as a model and technology for bringing tuple-based coordination to the Edge.

Different formalisms for defeasible reasoning have been used to represent legal knowledge and to reason with it. In this work, we provide an overview of the following logic-based approaches to defeasible reasoning: Defeasible Logic, Answer Set Programming, ABA+, ASPIC+, and DeLP. We compare features of these approaches from three perspectives: the logical model (knowledge representation), the method (computational mechanisms), and the technology (available software). On this basis, we identify and apply criteria for assessing their suitability for legal applications. We discuss the different approaches through a legal running example.

Distributing situated intelligence in Cyber-Physical Systems (CPS) to realise the vision of Internet of Intelligent Things (IoIT) raises issues of efficiency and scalability-in particular when dealing with huge numbers of physical objects. Such issues do not just regard the application or service logic and runtime, but also impact on the software development process. Moving from the notion of Logic Programming as a Service (LPaaS) - a re-interpretation of distributed logic programming tailored to the IoT era - in this paper we describe how its architecture and development process deals with the aforementioned issues from a software engineering standpoint, by discussing the design, development practices, and delivery means of the LPaaS technology.

Blockchain technologies are rapidly gaining attention in the multi-agent systems (MAS) community to face critical issues such as trust, secured communications, and data consistency. In particular, the notion of smart contract can be exploited to deploy trustworthy computations automatically executed by the network in a consistent way. MAS coordination - modelling and engineering of agents interaction in a MAS - thus represents an appealing application field for smart contracts, potentially enabling fully-decentralised, trustworthy coordination. Along this line, we focus on the Ethereum blockchain technology, map it onto LINDA tuple-based coordination model, and discuss two proof-of-concept implementations of LINDA on Ethereum. We hence demonstrate conceptual and technical feasibility of blockchain-based coordination in MAS, while emphasising issues of applying the blockchain beyond accountability and identity management.

Abstract We introduce Spatial Tuples, an extension of the basic tuple‐based model for distributed multi‐agent system coordination where (a) tuples are conceptually placed in regions of the physical world and possibly move anchored to a mobile computational device, (b) the behaviour of standard Linda coordination primitives is extended so as to depend on the spatial properties of the coordinating agents, tuples, and the topology of space, and (c) the tuple space can be conceived as a virtual layer augmenting physical reality. Motivated by the needs of mobile augmented‐reality applications, Spatial Tuples explicitly aims at supporting space‐aware and space‐based coordination in agent‐based pervasive computing scenarios. This paper presents the coordination model, its formalization as a process algebra, a library of patterns of coordination it enables, and a discussion of application scenarios, challenges, and open issues for future works.

In order to enable logic programming to deal with the diversity of pervasive systems, where many heterogeneous, domain-specific computational models could benefit from the power of symbolic computation, we explore the expressive power of labelled systems. To this end, we define a new notion of trut h for logic programs extended with labelled variables interpreted in non-Herbrand domains—where, however, terms maintain their usual Herbrand interpretations. First, a model for labelled variables in logic programming is defined. Then, the fixpoint and the operational semantics are presented and their equivalence is formally proved. A meta-interpreter implementing the operational semantics is also introduced, followed by some case studies aimed at showing the effectiveness of our approach in selected scenarios.

Abstract New generations of distributed systems are opening novel perspectives for logic programming (LP): On the one hand, service-oriented architectures represent nowadays the standard approach for distributed systems engineering; on the other hand, pervasive systems mandate for situated intelligence. In this paper, we introduce the notion ofLogic Programming as a Service(LPaaS) as a means to address the needs of pervasive intelligent systems through logic engines exploited as a distributed service. First, we define the abstract architectural model by re-interpreting classical LP notions in the new context; then we elaborate on the nature of LP interpreted as a service by describing the basic LPaaS interface. Finally, we show how LPaaS works in practice by discussing its implementation in terms of distributed tuProlog engines, accounting for basic issues such as interoperability and configurability.

Event-based systems (EBS) are nowadays the most viable sources of technologies and solutions for large-scale distributed applications. On the other hand, multi-agent systems (MAS) apparently provide the most viable abstractions and coherent methods to deal with complex distributed systems, in particular when advanced features - such as mobility, autonomy, symbolic reasoning, knowledge management, situation recognition - are required. In this position paper we discuss how the core concepts of EBS and MAS can in principle be matched and integrated, providing a sound conceptual ground for a coherent discipline for the engineering of complex software systems.

Changes and adaptations are always necessary after the deployment of a multi-agent system (MAS), as well as of any other type of software systems. Some of these changes may be simply perfective and have local impact only. However, adaptive changes to meet new situations in the operational environment of the MAS may impact globally on the overall design. More specifically, those changes usually affect the organizational structure of the MAS. In this paper we analyze the issue of design change/adaptation in a MAS organization, and the specific problem of how to properly model/design a MAS so as to make it ready for adaptation. Special attention is paid to the Gaia methodology, whose suitability in dealing with adaptable MAS organizations is also discussed with the help of an illustrative application example.

This article introduces the special issue of ACM Transactions on Autonomous and Adaptive Systems devoted to research papers arising from the three Technical Forum Group meetings held in 2004 and 2005 that were organized and sponsored by the European FP6 Coordination Action AgentLink III.

The notions of coordination, collaboration and cooperation have originated a number of heterogeneous research lines in diverse scientific areas, both inside and outside Computer Science — along with a number of different acceptations for the terms as well. The basic and almost obvious correlation between these terms, however, has not yet led to a satisfactory scientific overall picture that could put them in the right perspective, emphasizing their commonalities and distinctions. Still, there is a great potential for cross-fertilization between the different strands of work. In particular, if the mutual relationship between these terms was understood and commonly accepted, it would be much easier to adapt and apply many innovative ideas developed by the different communities to the field of cooperative information systems. In this article, we outline a possible unified conceptual framework, by suitably re-interpreting findings from Activity Theory. There, a clear definition for both coordination and cooperation as collaborative activities can be found and used as a reference, which is centred around the notion of (coordination) artifact. Then, we gladly introduce four contributions, selected from the best papers of the 18th ACM Symposium on Applied Computing and suitably revised for the International Journal on Cooperative Information Systems, that show how some of the most recent results of the research on coordination can be fruitfully exploited and applied to the field of cooperative information systems.

Objects and agents are abstractions that exhibit points of similarity, but the development of agent-based software poses other challenges to software engineering since software agents are inherently more complex entities. In addition, a large-scale multi-agent system needs to satisfy multiple stringent requirements such as reliability, security, interoperability, scalability, reusability, and maintainability. This workshop brought together researchers and practitioners to discuss the current state and future direction of research in software engineering for large-scale multi-agent systems. A particular interest was to understand those issues in the agent technology that difficult and/or improve the production of large-scale distributed systems.

Agents and multiagent systems (MAS) have already proved suitable for dealing with the complexity of today B2B application scenarios. Application environments such as virtual enterprises (VE) and inter-organizational workflow management systems (WfMS) specially call for the integration of heterogeneous resources, services, and processes. In this article, we first recognize VE and workflow management as agent coordination problems, then discuss how objective coordination — that is, coordination from outside agents — can help to model VE and WfMS. Finally, we show the impact of an agent coordination infrastructure like [Formula: see text] on the engineering of highly dynamic VE and WfMS, by discussing a simple case study.

No abstract available.

Multi-agent system development calls for powerful and expressive coordination models and languages, as well as for an effective coordination technology. A good deal of the current research effort focuses on tuple-based coordination, exploiting its well-known advantages, such as agent uncoupling and associative access to information, and addressing its limitations in terms of flexibility and control capabilities. In particular, the behaviour of a Linda-like tuple space is fixed once and for all by the coordination model, and cannot be tailored to the specific application needs. Tuple centres are tuple spaces whose behaviour can be programmed by defining transactional reactions to the basic communication events, allowing coordination laws to be explicitly defined and embedded into the coordination medium. This paper presents the architecture of a run-time system for tuple-based coordination, where tuple centres work as an extensible kernel, around which multi-agent systems can be designed and deployed. After sketching the implementation, the paper shows the advantages that can be achieved from both the design and the performance viewpoints. Copyright © 1999 John Wiley & Sons, Ltd.

This paper discusses a multi-agent framework for the control of robot systems along with an integrated programming environment which takes advantage of it. An agent society is engaged in a competitive or cooperative interaction under a control metalevel while a particular agent represents the real-time robot machine. By virtue of the support of adequate tools, this schema well lends itself both to a general increase of robot programming capability and flexibility and to rapid prototyping of different architectural solutions. The overall programming environment is built upon the integration of a distributed, extended logic programming environment with a real-time RCCL-based robot server. It is also interfaced to a graphical simulator in which a sensory driven task can be given a preliminary validation. An example of application of the environment to a simple test case concludes the paper.>