Services for data provisioning can be extremely valuable. Organizations wrestle with deciding whether to “expose or not to expose” their data through Web APIs and, depending on the nature of their business, mull over potential benefits and drawbacks. This is because data can be exploited and misused in unanticipated ways. Rate restriction techniques are frequently used to manage data access and protect back-end computing resources. The best choice for the maximum rate at which information can be “safely” provided to clients will determine its effectiveness. An important situation involves formal service-level agreements that govern the quality of services provided by public governments and private businesses. These firms might need to decide on a rate cap for their Web APIs that prevents unauthorized clients from accurately calculating service levels while still enabling the development of valuable value-added services. In this paper, we propose a statistical model for this problem and a technique, based on sampling tools, following brute-force, binary search, and heuristic search methods, to select an appropriate rate limit, and demonstrate its validity through a case study involving a large Italian bus company.

In today’s industrial environment, digital transformation is crucial for companies aiming to remain competitive and embrace smart manufacturing practices. Within this context, Maturity Models (MMs) have become valuable instruments for guiding smart evolution and ongoing improvement. Nevertheless, the effectiveness and practical use of an MM largely rely on the presence of web-based tools that present evaluation outcomes in a clear and user-friendly manner, enabling effective self-assessment or third-party evaluation. This paper introduces SM-FCM, a novel interactive web-based application designed to facilitate the implementation of an MM utilizing Fuzzy Cognitive Maps (FCMs), which helps in representing complex causal relationships between concepts. The presented platform simplifies the process of configuring model inputs, clearly shows the results of the assessment, and enables users to simulate and compare multiple scenarios with ease.

The Industry 4.0 paradigm is profoundly transforming aerospace assembly, integration, and testing (AIT) processes by fostering the integration of advanced digital technologies. This paper first introduces a novel framework that incorporates a domain-specific language (DSL) for structuring and standardizing assembly instructions, enabling automation, real-time adaptability, and reduced human error. This is complemented by advanced capabilities such as object recognition, tracking, and real-time sensor integration for contextual awareness and dynamic guidance. We then demonstrate the integration of cyber-physical systems (CPSs) and extended reality (XR) into aerospace AIT workflows, providing immersive, real-time support for operators. The framework is validated through the assembly of a CubeSat, in which 40 aerospace experts were divided into control and XR-assisted groups. Comparative evaluation demonstrated improvements in terms of efficiency, assembly accuracy, and overall usability. The XR-assisted group completed tasks more rapidly and with fewer errors, particularly in complex operations. Participants also reported higher satisfaction, perceiving the system as more intuitive and engaging than traditional methods. These results confirm the potential of XR- and DSL-driven CPS frameworks to enhance both operational performance and user experience in aerospace AIT.

Ensuring fair access to public administrative procedures is a significant challenge in e-government, due to their inherent complexity and the lack of standardized formats for textual descriptions. While open data initiatives have improved accessibility and transparency, many citizens and public sector employees still struggle to effectively interpret and utilize procedural information. To address this issue, we propose AdmPModeler, a novel tool that leverages Large Language Models, LLMs, to automatically generate process models from legal texts describing administrative procedures. Our tool employs advanced prompt engineering techniques —including prompt chaining, chain-of-thought reasoning, role prompting, and LLMs-as-a-judge— to enhance the quality and reliability of process modeling. We validate AdmPModelerthrough a human evaluation conducted in collaboration with domain experts from the Italian Department for Public Administration. The assessment compares the structured procedures extracted by the tool with their original textual descriptions, demonstrating its effectiveness in formally and clearly representing administrative procedures in a human-readable structured format.

Integrating multiple (sub-)systems is essential to create advanced Information Systems (ISs). Difficulties mainly arise when integrating dynamic environments across the IS lifecycle. A traditional approach is a registry that provides the API documentation of the systems' endpoints. Large Language Models (LLMs) have shown to be capable of automatically creating system integrations (e.g., as service composition) based on this documentation but require concise input due to input token limitations, especially regarding comprehensive API descriptions. Currently, it is unknown how best to preprocess these API descriptions. Within this work, we (i) analyze the usage of Retrieval Augmented Generation (RAG) for endpoint discovery and the chunking, i.e., preprocessing, of OpenAPIs to reduce the input token length while preserving the most relevant information. To further reduce the input token length for the composition prompt and improve endpoint retrieval, we propose (ii) a Discovery Agent that only receives a summary of the most relevant endpoints and retrieves details on demand. We evaluate RAG for endpoint discovery using the RestBench benchmark, first, for the different chunking possibilities and parameters measuring the endpoint retrieval recall, precision, and F1 score. Then, we assess the Discovery Agent using the same test set. With our prototype, we demonstrate how to successfully employ RAG for endpoint discovery to reduce the token count. While revealing high values for recall, precision, and F1, further research is necessary to retrieve all requisite endpoints. Our experiments show that for preprocessing, LLM-based and format-specific approaches outperform naïve chunking methods. Relying on an agent further enhances these results as the agent splits the tasks into multiple fine granular subtasks, improving the overall RAG performance in the token count, precision, and F1 score.

In recent years, growing attention to the gut microbiota has led to increasing use of complex clinical reports, such as those provided by microbiome and multi-omic analyses. Irritable bowel syndrome (IBS) is a relevant area of application, characterized by a chronic functional disorder of the gastrointestinal tract. Having clear and easily interpretable diagnostic tools increases the capacity of diagnosis and helps patients to better understand the content of the reports, which are often static and rich in scientific data, making them complex to understand for both patients and non-specialist physicians. This work proposes the development of a dynamic platform for the visualization and interactive management of microbiome clinical reports. Starting from existing reports, a prototype web interface capable of presenting data in a modular, adaptable, and customizable way has been designed and proposed. The prototype also allows the generation of reports in PDF format, both in a dynamic version (with the possibility of filtering and exploring different sections) and in a static version (useful for printing or clinical sharing). The added value of this solution lies in readability and quick access to the information.

This paper introduces the Safe-Cities Chatbot, a novel conversational AI system designed to support urban security planning and emergency management, with a focus on complex cross-border scenarios. The system integrates Retrieval-Augmented Generation (RAG) techniques with domain-specific knowledge from the Security Vulnerability Assessment (SVA) framework, enabling context-aware assistance for security professionals. Unlike existing solutions, the chatbot was developed and evaluated using a real-world test Piazza Transalpina, a public square on the Italian-Slovenian border incorporating local regulatory constraints, spatial dynamics, and multi-agency coordination challenges. The chatbot supports three core functions: guiding users through structured security assessments, offering event-specific planning advice, and simulating emergency scenarios tailored to the location. We detail the systems architecture, its contextual reasoning strategies, and the simulation engine design. Our results demonstrate the potential of AIdriven conversational tools to enhance situational awareness, improve preparedness, and facilitate security planning across jurisdictions a growing need in the face of new urban threats and regulatory requirements such as the NIS2 directive.

Artificial intelligence has become integral to the life sciences, enabling large-scale data analysis and accelerating discovery. However, deploying these analytical capabilities depends on cloud infrastructures that are often complex to design, validate, and operate. Current approaches rely heavily on manual, time-consuming configuration by experts. This paper explores how artificial intelligence, specifically Large Language Models, can support not only genomic data analysis but also the automated creation of the underlying computational infrastructure. We propose a methodology that translates naturallanguage user stories and requirements into cloud-architectural specifications and containerized deployments, bridging the gap between user intent and executable infrastructure. Preliminary results indicate that LLM-driven synthesis can reduce human effort in defining container relationships and deployment logic, advancing automation in cloud-native bioinformatics.

Anomaly detection in many industrial productions, including electronic boards, is a critical aspect of quality control, demanding high precision and reliability. Traditional inspection workflows rely heavily on manual expertise, making them labor-intensive, error-prone, and difficult to scale. In this paper, we present an open-source novel framework that integrates Large Language Models (LLMs) with Retrieval-Augmented Generation (RAG) and multimodal analysis to enhance anomaly detection. Our approach consists of two core components: (i) an AI-powered assistant capable of extracting and interpreting information from structured and unstructured documentation, and (ii) a multimodal LLM-based system that fuses textual and visual data to identify defects in electronic board images. To evaluate the framework, we conduct an extensive assessment using three open-source LLMs, benchmarking their performance on IPC-A-610F – the industry standard for electronic board defect detection. Key metrics, including accuracy, faithfulness, and retrieval precision, are analyzed to determine the system’s effectiveness. Results indicate that LLMs significantly enhance information retrieval and defect detection, providing a scalable, efficient solution for industrial applications.

In recent decades, the industrial world has undergone a significant transformation through the inclusion of innovative technologies that enhance manufacturing processes. In this context, Machine Vision inspection systems play a key role in ensuring quality by identifying defects in production. Automated defect detection systems improve productivity by reducing manual interventions, which can be time-consuming and prone to errors. This paper presents DIE-VIS, a realworld implemented visual inspection system for detecting defects in cardboard box manufacturing using traditional Computer Vision techniques. We provide a comprehensive evaluation comparing it to the YOLOv8 state-of-the-art deep learning model, demonstrating how, in the specific application of cardboard manufacturing, customized solutions still offer fundamental advantages.

In the current industrial landscape, digitization is essential for the survival and growth of companies of all sizes. The adoption of new technologies and organizational structures is critical as companies navigate complex digital transformations. While Maturity Models (MMs) offer significant support in these endeavors, existing models often overlook the integration of Information Technology systems, leading to gaps in assessing digital maturity. This paper addresses these shortcomings by (1) proposing a Fuzzy Cognitive Maps based MM which adeptly captures the causal relationships among IT systems and enabling technologies, and (2) defining and testing a methodology for constructing such a model with inputs from both industry experts and academic researchers. The new MM provides a holistic assessment of smart manufacturing maturity and outlines clear, context-specific pathways for digital transformation. A key feature of the proposed model is its practical applicability, demonstrated through a detailed case study, which validates the MM by applying it to a real-world manufacturing setting, assessing current digital maturity levels, and simulating various improvement scenarios. The results illustrate significant enhancements in digital maturity, affirming the model's practical value. This evidence supports the MM's utility in guiding companies toward informed digital advancements, with potential adaptations for specific markets and needs.

This study proposes a data-driven approach to leveraging large-scale COVID-19 datasets to enhance the predictive modeling of disease spread in the early stages. We systematically evaluate three machine learning models—ARIMA, Prophet, and LSTM—using a comprehensive framework that incorporates time-series analysis, multivariate data integration, and a Multi-Criteria Decision Making (MCDM) technique to assess model performance. The study focuses on key features such as daily confirmed cases, geographic variations, and temporal trends, while considering data constraints and adaptability across different scenarios. Our findings reveal that LSTM and ARIMA consistently outperform Prophet, with LSTM achieving the highest predictive accuracy in most cases, particularly when trained on 20-week datasets. ARIMA, however, demonstrates superior stability and reliability across varying time frames, making it a robust choice for short-term forecasting. A direct comparative analysis with existing approaches highlights the strengths and limitations of each model, emphasizing the importance of region-specific data characteristics and training periods. The proposed methodology not only identifies optimal predictive strategies but also establishes a foundation for automating predictive analysis, enabling timely and data-driven decision-making for disease control and prevention. This research is validated using data from New Zealand and its major trading partners—China, Australia, the United States, Japan, and Germany—demonstrating its applicability across diverse contexts. The results contribute to the development of adaptive forecasting frameworks that can empower public health authorities to respond proactively to emerging health threats.

This paper focuses on the aspects of responsibility and ethics when using AI systems in the contexts of health and genomics. The present proposal also aims to address the important need to consider how the use of AI systems in the field of genomics affects human life and its implications. As AI systems increasingly participate in diagnostic and predictive decision-making, they introduce new challenges concerning moral agency, transparency, and human oversight. The discussion aims to examine how these technologies, while offering unprecedented analytical capabilities, simultaneously reshape the ethical landscape of medical practice and genomic research.

Last year witnessed a growing interest from the Business Process Management research community in analyzing activities carried out in sensorized environments using techniques originally intended for business processes. However, activities conducted in such scenarios differ significantly from typical processes in terms of repetitiveness and predictability. This raises the issue of assessing the suitability of state-of-the-art modeling formalisms and mining techniques to represent them, especially when humans are involved. In this paper, we present the results of a user study conducted with this specific goal. Specifically, we analyze the opinions of a group of experts regarding different representation formalisms and mining algorithms, drawing conclusions about the usefulness of such models in smart environments.

In recent decades, manufacturing practices have undergone a significant transformation, with the integration of computers and automation playing a central role. Concurrently, there has been a growing interest in utilizing intelligent techniques to effectively manage manufacturing processes. These processes entail the seamless integration of various activities across the supply chain. Given the diverse range of actors in a supply chain, each one with distinct characteristics such as cost, quality, and probability of failure, task assignment becomes a crucial challenge. In such a complex scenario, manual decision-making becomes impractical, necessitating the adoption of automated techniques to effectively address these challenges in a resilient and adaptive manner. This article proposes a service-oriented approach to model each manufacturing actor within the supply chain. Furthermore, it categorizes automated synthesis approaches for smart manufacturing on the basis of i) the characteristics of each actor, which are retrieved by their Industrial API, and ii) the goal(s) of the manufacturing process. Finally, the article evaluates three distinct approaches that implement automated synthesis techniques for composing services and generating operational plans.

Nowadays, efficient management of energy consumption is crucial for the sustainability of our cities, and overall of our planet. Approaches investigated so far, mostly adopt complex approaches, often based on deep learning, which have an important footprint. This study focuses on the importance of using simpler methods to predict energy consumption in smart buildings, emphasizing a methodological approach that prioritizes simplicity, transparency, and computational efficiency, especially when data is scarce. It emphasizes that even the prediction of energy consumption at the scale of a building, which is sometimes ignored due to computational complexity, is feasible and can make a big difference. By using simple analytical models combined with outlier detection, this research contributes to the field by showing how we can still gain valuable insights with limited data. Therefore, this study provides a practical and scalable way to improve energy efficiency and sustainability in buildings, which has a significant contribution to energy management practices.

During the last years, a number of studies have experimented with applying process mining (PM) techniques to smart spaces data. The general goal has been to automatically model human routines as if they were business processes. However, applying process-oriented techniques to smart spaces data comes with its own set of challenges. This paper surveys existing approaches that apply PM to smart spaces and analyses how they deal with the following challenges identified in the literature: choosing a modelling formalism for human behaviour; bridging the abstraction gap between sensor and event logs; and segmenting logs in traces. The added value of this article lies in providing the research community with a common ground for some important challenges that exist in this field and their respective solutions, and to assist further research efforts by outlining opportunities for future work.

Cultural Heritage has become a considerable active area of application of linked data and semantic web technologies. Visualizing and exploring the linked data produced to enhance cultural heritage is one of the significant challenges in the research field of Digital Humanities. Therefore, in this survey, we describe systems developed by the semantic web community in the Web of Linked Data context. We compared twenty-eight interfaces that show both about generals that digital library contents. We classified them by interaction paradigm, type of information displayed and complexity reduction strategies used.

The study of ancient documents provides a glimpse into our past. However, the low image quality and intricate details commonly found in these documents present significant challenges for accurate object detection. The objective of this research is to enhance object detection in ancient documents by reducing false positives and improving precision. To achieve this, we propose a method that involves the creation of synthetic datasets through computational mediation, along with the integration of visual feature extraction into the object detection process. Our approach includes associating objects with their component parts and introducing a visual feature map to enable the model to discern between different symbols and document elements. Through our experiments, we demonstrate that improved object detection has a profound impact on the field of Paleography, enabling in-depth analysis and fostering a greater understanding of these valuable historical artifacts.