In the evolving landscape of artificial intelligence, ensuring the robustness and explainability of machine learning models is valuable. This study presents an innovative method based on the Rough Set Theory and Principles of Justified Granularity to enhance the explainability of text-based classifiers, specifically in style-based news bias classification. The method helps understand why a classifier can be deceived with an Adversarial Attack. It leverages two levels of insight. The first level is independent of the specific classifier and consists of generating rules from a boundary region built with Rough Sets Theory starting from train data. The second level considers the behavior of a specific machine learning model in classifying manipulated observations and, starting from the classification results, constructs information granules of true positives and false negatives. These granules are representative of observations that deceived a classifier. By comparing boundary rules with information granules, it is possible to acquire actionable knowledge that is useful for making decisions on making a machine learning model more resilient. Results are evaluated with real data containing biased news. The success rate of adversarial examples generated using LLM to test classifiers on borderline cases, where minor textual changes cause false negatives, ranges from 45% to 68%.

This paper presents a novel approach for predicting and explaining instances of Information Disorder. The paper reports two significant findings: i) the use of structures of opposition to describe relationships between instances of Information Disorder, and ii) the development of an explainable prediction method that combines Fuzzy Rough Sets and TOPSIS with these structures. The findings have the potential to assist analysts and decision-makers in gaining a deeper understanding of the phenomenon of Information Disorder. The results are based on real data and demonstrate promising applications for future research.

The paper presents and evaluates an approach based on Rough Set Theory, and some variants and extensions of this theory, to analyze phenomena related to Information Disorder. The main concepts and constructs of Rough Set Theory, such as lower and upper approximations of a target set, indiscernibility and neighborhood binary relations, are used to model and reason on groups of social media users and sets of information that circulate in the social media. Information theoretic measures, such as roughness and entropy, are used to evaluate two concepts, Complexity and Milestone, that have been borrowed by system theory and contextualized for Information Disorder. The novelty of the results presented in this paper relates to the adoption of Rough Set Theory constructs and operators in this new and unexplored field of investigation and, specifically, to model key elements of Information Disorder, such as the message and the interpreters, and reason on the evolutionary dynamics of these elements. The added value of using these measures is an increase in the ability to interpret the effects of Information Disorder, due to the circulation of news, as the ratio between the cardinality of lower and upper approximations of a Rough Set, cardinality variations of parts, increase in their fragmentation or cohesion. Such improved interpretative ability can be beneficial to social media analysts and providers. Four algorithms based on Rough Set Theory and some variants or extensions are used to evaluate the results in a case study built with real data used to contrast disinformation for COVID-19. The achieved results allow to understand the superiority of the approaches based on Fuzzy Rough Sets for the interpretation of our phenomenon.

Abstract The paper presents the results related to the definition and adoption of structures of opposition, built with set-theoretic measures, to evaluate emotional dynamics that arise during conversations on social media. Specifically, a graded hexagon of opposition is used to compare the emotional profiles of individuals involved in a dyadic conversation. Set-theoretic measures, based on fuzzy logic, are used to construct the hexagon whose analysis allows us to understand the tendency of the conversation toward empathy or lack of empathy. The results can be useful in the context of the current trend of social media sensing and, in particular, to support social media providers in receiving early warnings related to the analysis of emotional dynamics that could lead to or degenerate into information disorder. The results have been evaluated with conversations extracted from the Empathetic Dialogue dataset.

This paper presents a comprehensive model for representing and reasoning on situations to support decision makers in Intelligence analysis activities. The main result presented in the paper stems from a work of refinement and abstraction of previous results of the authors related to the use of Situation Awareness and Granular Computing for the development of analysis methods and techniques to support Intelligence. This work made it possible to derive the characteristics of the model from previous case studies and applications with real data, and to link the reasoning techniques to concrete approaches used by intelligence analysts such as, for example, the Structured Analytic Techniques. The model allows to represent an operational situation according to three complementary perspectives: descriptive, relational and behavioral. These three perspectives are instantiated on the basis of the principles and methods of Granular Computing, mainly based on the theories of fuzzy and rough sets, and with the help of further structures such as graphs. As regards the reasoning on the situations thus represented, the paper presents four methods with related case studies and applications validated on real data.

This article presents a new interactive method to analyze and assess hypotheses, and its application to terrorism events. The method combines probability, fuzzy, and rough set theories and supports decision makers and analysts of counterterrorism in the analysis of intelligence information by using behavioral models of known terrorist groups. Starting from intelligence information about possible attack patterns, the proposed method uses two parameters allowing derivation and analysis of a wide range of hypotheses, and their assessment on the basis of different support levels of evidence. The evaluation of results has been done on real data relating to five years (2012-2016) of terrorist activities extracted from the Global Terrorism Database.

A great impetus for the study of resilience in critical infrastructures (CIs) is found in the large number of initiatives and international research programmes from U.S., EU, and Asia. Politicians, decision makers, and citizens are now aware of the drastic consequences that can have the cascading effects of an adverse event in these large scale infrastructures. However, the study of resilience in CIs is challenging for several reasons, among which their large scale and interdependencies. We have to consider also that adverse events, e.g., attacks, natural hazards, or man-made disasters, suddenly occur and evolve rapidly, giving us little time to take decisions and react to them. Approximate reasoning and rapid decision making have to be considered requirements for resilience analysis of CIs. The main result presented in this paper relates to a systemic integration of granular computing (GrC) and resilience analysis for CIs. Each phase of our approach presents distinctive aspects but, overall, we argue the merit of this paper consists in the originality of the study, being this the first work that combines GrC and resilience analysis of CIs. This paper reports an illustrative example that shows how to apply our results, and a discussion on the necessary contextualizations and extensions of the GrC results to be better adapted for CIs resilience.

We present a methodology to support urban planners and decision makers in obtaining a good awareness of how city assets (points of interest) are perceived by a community, and on the impact and influence that this collective perception can have on other city assets and city issues such as mobility, environment, security. The methodology employees Fuzzy Cognitive Maps and Fuzzy sets. Fuzzy Cognitive Maps are used to model the relationships between elements of mental representations that different communities have with regards to city issues. The concept of signature as relation between two fuzzy sets is adopted, in analogy to what proposed by Yager and Reformat [1], to characterize a point of interest. Different signatures are subsequently grouped to characterize an area and adopted, in combination with sentiment analysis, to derive a measure of collective perception on the quality of the area. This measure is used to activate some qualitative concept of a Fuzzy Cognitive Map and perform what-if analysis. The methodology has been applied to a sample of three POIs (representing three attractions of the city of Salerno) by using data gathered from the Web and involving some real citizens. Our preliminary results are encouraging with regards to the possibilities offered by our approach of enforcing city decision makers with a good awareness on how changes in the perception of quality of urban areas can influence other city related issues.

We present an application of Granular Computing and Three-way decisions to intelligence analysis. In particular we extend the Analysis of Competing Hypotheses with an additional perspective devoted to support analysts in reasoning with groups of hypotheses that can be equivalent on the basis of partial and incomplete evidence, and in classifying these groups of hypotheses with respect to a decisional attribute of interest for the analyst, such as dangerous or safe. Creating and reasoning with granules and multi-level granular structures give to our approach an added value when dealing with a large number of evidence and hypotheses. Three-way decision making offers the possibility of a rapid understanding of how granules of hypotheses approximate a class of dangerous hypotheses, with clear benefits when analysts have to take decision on classifying a group of hypotheses or setting a proper level of attention to group of equivalent hypotheses.

In this paper, we present our results related to the definition of a methodology that combines augmented reality (AR) with semantic techniques for the creation of digital stories associated with museum exhibitions. In contrast to traditional AR approaches, we augment real-world elements by supplementing contents of a museum exhibition with additional inputs that provide new and different meanings. In this way we augment a cultural resource with respect to both its presentation and meaning. The methodology is framed in the cultural re-mediation theory and is grounded on a set of ontologies aimed at modelling a cultural resource and correlating it with external multimedia objects and resources. To provide an easy tool for the creation of museum narratives, the methodology makes use of a set of recognised practices widely adopted by museum curators that have been formalised through inference rules. The defined methodology has been experimented in a scenario related to Flemish paintings to validate the augmentation of cultural objects with two different approaches, the first basing on similarities and the second on dissimilarities.

We present our preliminary results in the definition of a model and knowledge based techniques to support creativity by establishing consonance relationships between a worker and the organization. The model is based on the Viable Systems Approach (VSA) and links this theory with Creativity and Open innovation via the definition of proper consonance and resonance relationships. VSA is an interdisciplinary approach grounded on systems thinking and resource-based theory, and focused on methodologies to govern relations among supra-systems and subsystems. Leveraging on the systems perspective of the creativity proposed by Csikszentmihalyi, we defined a way to analyse and understand how the variety introduced by workers in organization can lead to novel and useful products and services. We present also our preliminary results on how a well-defined set of knowledge based methodologies and techniques, resulting from the ARISTOTELE research project, can be applied to this purpose in the context of a Gene lore model. We believe our proposal, i.e. the application of VSA methods combined with knowledge based techniques, can lead to substantial advantages in several areas of Computational Creativity.

The wealth of modern enterprises has progressively shifted from tangible assets (capital, resources,) into intangible ones (knowledge, reputation, skills management, innovation processes,). Intangibles are closely related to the natural interactions normally occurring among work practices. This is where ideas, innovation, learning, knowledge, social cohesion, and other diverse intangibles synergistically contribute to performance, competition differentiators and value creation. In order to make these intangibles productive for the organization, we have to define suitable models for their correct representation in real contexts, as well as tools for their proper management in the workplace environment in a transparent way. All these aspects are the foundation of the ARISTOTELE research project. In this paper, we address two issues: i) the use of semantic technologies for modeling and cross relating relevant organizational assets, namely knowledge, competency, worker and learning, ii) how to design a software architecture for managing these models through the integration of ad-hoc developed tools and commercial off-the-shelf platforms.

The paper presents the main findings of the ELeGI project, namely its learning model and software architecture to support the creation and execution of complex learning processes.The learning model defined in ELeGI promotes and supports a learning paradigm centred on knowledge construction using experiential based and collaborative learning approaches in a contextualised, personalised and ubiquitous way.The software architecture has been designed and developed taking into account the learning model for the personalisation of complex learning experiences.In order to validate our results, the paper presents and describes a case study relating to the implementation of a Unit of Learning for explanation of the Torricelli's law, and its execution on top of the Service Oriented Architecture.

In this paper, starting from the limitations and constrains of traditional human learning approaches, we outline new suitable approaches to education and training in future knowledge based society. In our vision, learning and teaching are no longer standalone activities but complex, conversational and experiential-based processes implying collaboration, direct experience, mutual trust and shared interests. We identify characteristics of the environments suitable for these processes, and we compare different enabling technology infrastructures in order to justify why the semantic grid for human learning, that is a particular enhanced instance of the traditional semantic grid, is the most appropriate infrastructure to build our vision on. Finally, we present a realistic learning scenario as a case study, proving the effectiveness of our innovative learning approaches for future education and training.