Traditional Intrusion Detection Systems (IDS) are increasingly challenged by the distributed, heterogeneous, and rapidly evolving threat landscape in Internet of Things (IoT) and Industrial IoT (IIoT) environments. Blockchain has been explored as a promising foundation for decentralized and trustworthy security mechanisms; however, the existing literature remains fragmented and lacks a clear organizing lens for comparing design choices and evaluation practices. To address this, this paper presents a problem-driven survey of blockchain-enabled IDS for IoT and IIoT. We organize prior work into four integration paradigms, Trusted Rule, ML, DL, and FL—and relate each paradigm to the recurring design tensions it primarily targets. We further distill three fundamental tensions that frequently shape system design, including distributed architectures vs. centralized security management, collaborative information sharing vs. privacy preservation, and real-time detection requirements vs. resource-constrained devices. In addition, we summarize representative frameworks by consolidating datasets, threat models, and reported performance-related metrics, and we discuss common limitations that hinder cross-paper comparability. Finally, we outline a roadmap toward more standardized benchmarking, suggesting candidate evaluation criteria and blockchain-specific KPIs to encourage more transparent and comparable reporting. Overall, this survey aims to provide a structured lens for navigating the design space of blockchain-enabled IDS and to highlight open challenges for future research.

Large Language Models (LLMs) show strong performance in natural language processing tasks, but their application in the financial domain is limited. Current methods rely on large datasets and manual prompt engineering, resulting in high data demands, long inference times, and frequent hallucinations. To address these limitations, we propose a novel self-supervised prompt optimization framework tailored for the financial domain. Our approach involves training a critical model that evaluates and ranks generated outputs using both good and bad answers generated from various revised prompts. Experiments on a large Chinese financial corpus show that our framework significantly improves performance on tasks such as summarization and event-based question answering, as evidenced by higher scores on both automated metrics like ROUGE, BLEU, and BERTScore, and also through human evaluations. These results validate the effectiveness of our method in reducing hallucinations and improving the quality of financial text generation.

With the increasing prevalence of online communication and automated services, dialogue summarization technology plays a vital role in meeting minutes, customer service, and online Q&A scenarios. However, existing methods often suffer from insufficient flexibility in topic segmentation, low efficiency in semantic information transfer, and limited role adaptability. To address these challenges, we propose DABART, a dynamic semantic optimization framework. The framework employs a dynamic semantic topic segmentation mechanism to adaptively segment topics based on the distribution characteristics of sentence embeddings within dialogues, effectively identifying key information while overcoming the limitations of fixed-parameter methods in complex dialogue scenarios. Additionally, a dynamic semantic bridging module integrates semantic and positional information, further enhancing the coherence and consistency of dialogue summarization. Experimental results demonstrate that DABART achieves superior performance on widely-used benchmarks such as SAMSum and CSDS. Notably, it surpasses state-of-the-art open-source models on the CSDS dataset in ROUGE and BERTScore metrics, while achieving more balanced and accurate role-oriented summarization. Extensive experimental analyses further validate the robustness and applicability of the DABART across diverse dialogue scenarios.

Unlocking the potential of Industry 5.0 hinges on robust cybersecurity measures. This new Industrial Revolution prioritises human-centric values while addressing pressing soci-etal issues such as resource conservation, climate change, and social stability. Recognising the heightened risk of cyberattacks due to the new enabling technologies in Industry 5.0, this paper analyses potential threats and corresponding countermeasures. Furthermore, it evaluates the existing industrial implementation frameworks, which reveals their inadequacy in ensuring a secure transition from Industry 4.0 to Industry 5.0. Consequently, the paper underscores the necessity of developing a new framework centred on cybersecurity to facilitate organisations' secure adoption of Industry 5.0 principles. The creation of such a framework is emphasised as a necessity for organisations.

Security Information and Event Management (SIEM) solutions collect events from the IT infrastructure and concentrate information from the various components in a single place, allowing the detection of anomalous situations and attacks, and helping to protect confidential data. But real-world network environments may be complex and heterogeneous (e.g., in terms of devices, applications, and operating systems), and the attack surface can be vast, which makes increases the amount that a SIEM solution must collect and analyze. The General Data Protection Regulation (GDPR) has increased the level of complexity in such context, as organizations must ensure the monitoring of access to personal data and various levels of security in their infrastructure. In this work, we deal with the implementation of an open-source SIEM solution that incorporates technical measures for the protection and control of personal data, ensuring compliance with the GDPR. We identify the main functionalities and describe a solution based on the Elastic Stack and additional open-source external tools. To validate our proposals, we implemented a prototype of our solution in a real-world environment. We simulated internal and external attacks that show the solution capacity to deal in real-time with the detection of threats and incidents. We also evaluated the performance and resource consumption of personal data pseudonymization processes. Obtained results show our solution presents good performance and scalability.

Internet of Things (IoT) devices have become day-to-day technologies. They collect and share a large amount of data, including private data, and are an attractive target of potential attackers. On the other hand, machine learning has been used in several contexts to analyze and classify large volumes of data. Hence, using machine learning to classify network traffic data and identify anomalous traffic and potential attacks promises. In this work, we use deep and traditional machine learning to identify anomalous traffic in the IoT-23 dataset, which contains network traffic from real-world equipment. We apply feature selection and encoding techniques and expand the types of networks evaluated to improve existing results from the literature. We compare the performance of algorithms in binary classification, which separates normal from anomalous traffic, and in multiclass classification, which aims to identify the type of attack.