Computer Standards & Interfaces最新文献_第4页

Integrating IoT security practices into a risk-based framework for small and medium enterprises (SMEs) 将物联网安全实践整合到中小型企业（SMEs）的基于风险的框架中

IF 3.1 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computer Standards & Interfaces

Pub Date : 2025-11-26 DOI: 10.1016/j.csi.2025.104099

Samer Aoudi, Hussain Al-Aqrabi

The growing integration of Internet of Things (IoT) technologies within Small and Medium Enterprises (SMEs) has introduced new operational efficiencies while simultaneously expanding the cybersecurity threat landscape. However, most SMEs lack the resources, technical expertise, and institutional maturity required to adopt existing security frameworks, which are often designed with large enterprises in mind. This paper proposes a risk-based framework specifically developed to help SMEs identify, assess, and mitigate IoT-related security risks in a structured and scalable manner. The framework integrates key components such as asset classification, STRIDE-based threat modeling, CVSS-driven vulnerability assessment, and dynamic risk prioritization through Bayesian inference. Emphasis is placed on cost-effective mitigation strategies that are feasible within SME resource constraints and aligned with regulatory requirements. The framework was validated through a real-world case study involving a digitally enabled retail SME. Results demonstrate tangible improvements in vulnerability management, security control implementation, and organizational readiness. Additionally, qualitative feedback from stakeholders highlights the framework’s usability, adaptability, and minimal disruption to operations. This research bridges a critical gap in the current literature by contextualizing established cybersecurity methodologies for the SME sector and providing a practical toolset for managing IoT risks. The proposed framework offers SMEs a viable path toward improving cybersecurity resilience in increasingly connected business environments.

物联网（IoT）技术在中小企业（SMEs）中的日益融合，在带来新的运营效率的同时，也扩大了网络安全威胁的范围。然而，大多数中小企业缺乏采用现有安全框架所需的资源、技术专长和制度成熟度，这些框架通常是为大型企业设计的。本文提出了一个基于风险的框架，专门用于帮助中小企业以结构化和可扩展的方式识别、评估和减轻物联网相关的安全风险。该框架集成了资产分类、基于stride的威胁建模、cvss驱动的漏洞评估以及通过贝叶斯推理进行的动态风险优先级排序等关键组件。重点是在中小企业资源限制范围内可行并符合监管要求的具有成本效益的缓解战略。该框架通过涉及数字化零售中小企业的实际案例研究进行了验证。结果显示了在漏洞管理、安全控制实现和组织准备方面的切实改进。此外，来自涉众的定性反馈强调了框架的可用性、适应性和对操作的最小干扰。本研究通过将中小企业部门的既定网络安全方法置于背景中，并提供管理物联网风险的实用工具集，弥合了当前文献中的关键空白。拟议的框架为中小企业在日益互联的商业环境中提高网络安全弹性提供了一条可行的途径。

{"title":"Integrating IoT security practices into a risk-based framework for small and medium enterprises (SMEs)","authors":"Samer Aoudi, Hussain Al-Aqrabi","doi":"10.1016/j.csi.2025.104099","DOIUrl":"10.1016/j.csi.2025.104099","url":null,"abstract":"<div><div>The growing integration of Internet of Things (IoT) technologies within Small and Medium Enterprises (SMEs) has introduced new operational efficiencies while simultaneously expanding the cybersecurity threat landscape. However, most SMEs lack the resources, technical expertise, and institutional maturity required to adopt existing security frameworks, which are often designed with large enterprises in mind. This paper proposes a risk-based framework specifically developed to help SMEs identify, assess, and mitigate IoT-related security risks in a structured and scalable manner. The framework integrates key components such as asset classification, STRIDE-based threat modeling, CVSS-driven vulnerability assessment, and dynamic risk prioritization through Bayesian inference. Emphasis is placed on cost-effective mitigation strategies that are feasible within SME resource constraints and aligned with regulatory requirements. The framework was validated through a real-world case study involving a digitally enabled retail SME. Results demonstrate tangible improvements in vulnerability management, security control implementation, and organizational readiness. Additionally, qualitative feedback from stakeholders highlights the framework’s usability, adaptability, and minimal disruption to operations. This research bridges a critical gap in the current literature by contextualizing established cybersecurity methodologies for the SME sector and providing a practical toolset for managing IoT risks. The proposed framework offers SMEs a viable path toward improving cybersecurity resilience in increasingly connected business environments.</div></div>","PeriodicalId":50635,"journal":{"name":"Computer Standards & Interfaces","volume":"97 ","pages":"Article 104099"},"PeriodicalIF":3.1,"publicationDate":"2025-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145694121","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Fully decentralized period k-times anonymous authentication with access criteria 具有访问标准的完全分散的k次匿名身份验证

IF 3.1 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computer Standards & Interfaces

Pub Date : 2025-11-19 DOI: 10.1016/j.csi.2025.104097

Hongyan Di , Yinghui Zhang , Ziqi Zhang , Yibo Pang , Rui Guo , Yangguang Tian

The explosive growth of Internet user devices highlights the strong and urgent need for digital identity infrastructure. However, the existing decentralized identity schemes are still not fully decentralized, and there is still a contradiction between publicly auditable credentials and maintaining anonymity. Therefore, using advanced cryptographic techniques such as signature proof of knowledge, Pedersen commitment, and Merkle tree, this paper propose a fully decentralized period k-times anonymous authentication with access criteria. The scheme allows user credentials to be publicly audited, users can manage their identity independently, and the verifier can not only verify the user’s identity, but also implement access control. The issuer does not need to hold a key or maintain a list, and it can still authenticate even after the trusted center is attacked, and only three zero-knowledge proofs are needed for registration and verification. The security analysis indicates that this scheme satisfies unforgeability, anonymity, unlinkability and attribute privacy. Performance evaluation shows significant improvements in both computational and communication efficiency over existing schemes.

互联网用户设备的爆炸式增长凸显了对数字身份基础设施的强烈而迫切的需求。然而，现有的去中心化身份方案仍然没有完全去中心化，公开可审计的凭证和保持匿名之间仍然存在矛盾。因此，本文利用知识签名证明、Pedersen承诺和Merkle树等先进的密码学技术，提出了一种具有访问标准的完全分散的k次匿名身份验证方法。该方案允许对用户凭证进行公开审计，用户可以独立管理自己的身份，验证者不仅可以验证用户的身份，还可以实现访问控制。发行者不需要持有密钥，也不需要维护列表，即使在可信中心受到攻击后，发行者仍然可以进行身份验证，注册和验证只需要三个零知识证明。安全性分析表明，该方案满足不可伪造性、匿名性、不可链接性和属性隐私性。性能评估表明，与现有方案相比，计算效率和通信效率都有显著提高。

{"title":"Fully decentralized period k-times anonymous authentication with access criteria","authors":"Hongyan Di , Yinghui Zhang , Ziqi Zhang , Yibo Pang , Rui Guo , Yangguang Tian","doi":"10.1016/j.csi.2025.104097","DOIUrl":"10.1016/j.csi.2025.104097","url":null,"abstract":"<div><div>The explosive growth of Internet user devices highlights the strong and urgent need for digital identity infrastructure. However, the existing decentralized identity schemes are still not fully decentralized, and there is still a contradiction between publicly auditable credentials and maintaining anonymity. Therefore, using advanced cryptographic techniques such as signature proof of knowledge, Pedersen commitment, and Merkle tree, this paper propose a fully decentralized period <em>k</em>-times anonymous authentication with access criteria. The scheme allows user credentials to be publicly audited, users can manage their identity independently, and the verifier can not only verify the user’s identity, but also implement access control. The issuer does not need to hold a key or maintain a list, and it can still authenticate even after the trusted center is attacked, and only three zero-knowledge proofs are needed for registration and verification. The security analysis indicates that this scheme satisfies unforgeability, anonymity, unlinkability and attribute privacy. Performance evaluation shows significant improvements in both computational and communication efficiency over existing schemes.</div></div>","PeriodicalId":50635,"journal":{"name":"Computer Standards & Interfaces","volume":"97 ","pages":"Article 104097"},"PeriodicalIF":3.1,"publicationDate":"2025-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145625188","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An autonomous deep reinforcement learning-based approach for memory configuration in serverless computing 无服务器计算中基于自主深度强化学习的内存配置方法

IF 3.1 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computer Standards & Interfaces

Pub Date : 2025-11-19 DOI: 10.1016/j.csi.2025.104098

Zahra Shojaee Rad, Mostafa Ghobaei-Arani, Reza Ahsan

Serverless computing has become very popular in recent years due to its cost savings and flexibility. Serverless computing is a cloud computing model that allows developers to create and deploy code without having to manage the infrastructure. It has been embraced due to its scalability, cost savings, and ease of use. However, memory configuration is one of the important challenges in serverless computing due to the transient nature of serverless functions, which are stateless and ephemeral. In this paper, we propose an autonomous approach using deep reinforcement learning and a reward mechanism for memory configuration called Auto Opt Mem. In the Auto Opt Mem mechanism, the system learns to allocate memory resources to serverless functions in a way that balances overall performance and minimizes wastage of resources. Finally, we validate the effectiveness of our solution, the findings revealed that Auto Opt Mem mechanism enhances resource utilization, reduces operation cost and latency, and improves quality of service (QoS). Our experiments demonstrate that Auto Opt Mem mechanism achieves 16.8 % lower latency compared to static allocation, 11.8 % cost reduction, and 6.8 % improve in QoS, resource utilization, and efficient memory allocation compared with base-line methods.

近年来，无服务器计算由于其节省成本和灵活性而变得非常流行。无服务器计算是一种云计算模型，它允许开发人员创建和部署代码，而无需管理基础设施。它因其可伸缩性、成本节约和易用性而受到欢迎。然而，内存配置是无服务器计算中的重要挑战之一，因为无服务器功能具有瞬时特性，它们是无状态和短暂的。在本文中，我们提出了一种使用深度强化学习和记忆配置奖励机制的自主方法，称为Auto Opt Mem。在Auto Opt内存机制中，系统学习以平衡整体性能和最小化资源浪费的方式将内存资源分配给无服务器功能。最后，我们验证了该方案的有效性，结果表明，自动选择内存机制提高了资源利用率，降低了运行成本和延迟，提高了服务质量（QoS）。我们的实验表明，与静态分配相比，自动选择内存机制的延迟降低了16.8%，成本降低了11.8%，在QoS、资源利用率和有效内存分配方面比基线方法提高了6.8%。

{"title":"An autonomous deep reinforcement learning-based approach for memory configuration in serverless computing","authors":"Zahra Shojaee Rad, Mostafa Ghobaei-Arani, Reza Ahsan","doi":"10.1016/j.csi.2025.104098","DOIUrl":"10.1016/j.csi.2025.104098","url":null,"abstract":"<div><div>Serverless computing has become very popular in recent years due to its cost savings and flexibility. Serverless computing is a cloud computing model that allows developers to create and deploy code without having to manage the infrastructure. It has been embraced due to its scalability, cost savings, and ease of use. However, memory configuration is one of the important challenges in serverless computing due to the transient nature of serverless functions, which are stateless and ephemeral. In this paper, we propose an autonomous approach using deep reinforcement learning and a reward mechanism for memory configuration called Auto Opt Mem. In the Auto Opt Mem mechanism, the system learns to allocate memory resources to serverless functions in a way that balances overall performance and minimizes wastage of resources. Finally, we validate the effectiveness of our solution, the findings revealed that Auto Opt Mem mechanism enhances resource utilization, reduces operation cost and latency, and improves quality of service (QoS). Our experiments demonstrate that Auto Opt Mem mechanism achieves 16.8 % lower latency compared to static allocation, 11.8 % cost reduction, and 6.8 % improve in QoS, resource utilization, and efficient memory allocation compared with base-line methods.</div></div>","PeriodicalId":50635,"journal":{"name":"Computer Standards & Interfaces","volume":"97 ","pages":"Article 104098"},"PeriodicalIF":3.1,"publicationDate":"2025-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145580388","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Traceable signatures from lattices 来自格子的可追踪签名

IF 3.1 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computer Standards & Interfaces

Pub Date : 2025-11-13 DOI: 10.1016/j.csi.2025.104092

Nam Tran , Khoa Nguyen , Dongxi Liu , Josef Pieprzyk , Willy Susilo

Traceable signatures (Kiayas et al., 2004) is an anonymous digital signature system that extends the tracing power of opening authority in group signatures. There are many known constructions of traceable signatures but all are based on number-theoretical/pairing assumptions. For such reason, they may not be secure in the presence of quantum computers. This work revisits the notion of traceable signatures and presents a lattice-based construction provably secure in the quantum random oracle model (QROM).

可跟踪签名（Kiayas et al., 2004）是一种匿名数字签名系统，它扩展了群签名中开放权限的跟踪能力。有许多已知的可追踪签名结构，但都是基于数字理论/配对假设。因此，在量子计算机的存在下，它们可能不安全。这项工作重新审视了可追踪签名的概念，并提出了一个基于格子的构造，可以证明在量子随机预言模型（QROM）中是安全的。

引用次数: 0

Enhancing decision-making in Industry 5.0 through adaptive human–machine interfaces: A systematic literature review 通过自适应人机界面增强工业5.0中的决策：系统的文献综述

IF 3.1 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computer Standards & Interfaces

Pub Date : 2025-11-13 DOI: 10.1016/j.csi.2025.104091

Salima Meftah , M’hammed Sahnoun , Mourad Messaadia , Sidi-Mohamed Benslimane

In the dynamic landscape of Industry 5.0, adaptive human–machine interface (HMI) plays a pivotal role in shaping decision-making processes. This study constitutes a systematic literature review focusing on adaptive HMI in Industry 5.0, exploring their applications and implications within the decision-making context. The research objectives are structured around key questions, addressing the manifestation of adaptive HMI in the context of Industry 5.0, the evolutionary trajectory of HMI across industrial revolutions, relevant studies on adaptive HMI, the discernible impact of contextual use on HMI adaptation, emerging challenges of the new generation of HMI within Industry 5.0, and the integration of advanced technologies in adaptive HMI. Our findings identify key factors shaping decision-making in adaptive HMI, including acceptance and trust, transparency and reliability, privacy and security, human-centric design, long-term usability effects, accessibility, and decision support capabilities. We also highlight the role of enabling technologies—such as the Internet of Things (IoT), Internet of Humans (IoH), Virtual and Augmented Reality (VR/AR), Artificial Intelligence (AI), the Internet of Everything (IoE), and their synergistic integration in advancing adaptive HMI capabilities. This research provides valuable insights for designing and developing the next generation of adaptive HMI, specifically aimed at enhancing decision-making operators’ productivity and improving the quality of their decisions in smart manufacturing environments within Industry 5.0.

在工业5.0的动态环境中，自适应人机界面（HMI）在塑造决策过程中起着关键作用。本研究对工业5.0中的自适应人机界面进行了系统的文献综述，探讨了它们在决策环境中的应用和影响。研究目标围绕以下关键问题展开：工业5.0背景下自适应HMI的表现、HMI在工业革命中的演变轨迹、适应性HMI的相关研究、情境使用对HMI适应的明显影响、工业5.0中新一代HMI面临的新挑战，以及先进技术在适应性HMI中的整合。我们的研究结果确定了影响自适应人机界面决策的关键因素，包括接受度和信任、透明度和可靠性、隐私和安全性、以人为本的设计、长期可用性影响、可访问性和决策支持能力。我们还强调了使能技术的作用，如物联网（IoT）、人联网（IoH）、虚拟和增强现实（VR/AR）、人工智能（AI）、万物互联（IoE），以及它们在推进自适应人机界面能力方面的协同整合。本研究为设计和开发下一代自适应人机界面提供了有价值的见解，特别是针对工业5.0智能制造环境中提高决策操作员的生产力和决策质量。

{"title":"Enhancing decision-making in Industry 5.0 through adaptive human–machine interfaces: A systematic literature review","authors":"Salima Meftah , M’hammed Sahnoun , Mourad Messaadia , Sidi-Mohamed Benslimane","doi":"10.1016/j.csi.2025.104091","DOIUrl":"10.1016/j.csi.2025.104091","url":null,"abstract":"<div><div>In the dynamic landscape of Industry 5.0, adaptive human–machine interface (HMI) plays a pivotal role in shaping decision-making processes. This study constitutes a systematic literature review focusing on adaptive HMI in Industry 5.0, exploring their applications and implications within the decision-making context. The research objectives are structured around key questions, addressing the manifestation of adaptive HMI in the context of Industry 5.0, the evolutionary trajectory of HMI across industrial revolutions, relevant studies on adaptive HMI, the discernible impact of contextual use on HMI adaptation, emerging challenges of the new generation of HMI within Industry 5.0, and the integration of advanced technologies in adaptive HMI. Our findings identify key factors shaping decision-making in adaptive HMI, including acceptance and trust, transparency and reliability, privacy and security, human-centric design, long-term usability effects, accessibility, and decision support capabilities. We also highlight the role of enabling technologies—such as the Internet of Things (IoT), Internet of Humans (IoH), Virtual and Augmented Reality (VR/AR), Artificial Intelligence (AI), the Internet of Everything (IoE), and their synergistic integration in advancing adaptive HMI capabilities. This research provides valuable insights for designing and developing the next generation of adaptive HMI, specifically aimed at enhancing decision-making operators’ productivity and improving the quality of their decisions in smart manufacturing environments within Industry 5.0.</div></div>","PeriodicalId":50635,"journal":{"name":"Computer Standards & Interfaces","volume":"96 ","pages":"Article 104091"},"PeriodicalIF":3.1,"publicationDate":"2025-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145578671","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Graph-based interpretable dialogue sentiment analysis: A HybridBERT-LSTM framework with semantic interaction explainer 基于图的可解释对话情感分析：一个带有语义交互解释器的HybridBERT-LSTM框架

IF 3.1 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computer Standards & Interfaces

Pub Date : 2025-11-12 DOI: 10.1016/j.csi.2025.104086

Ercan Atagün , Günay Temür , Serdar Bi̇roğul

Conversational sentiment analysis in natural language processing faces substantial challenges due to intricate contextual semantics and temporal dependencies within multi-turn dialogues. We present a novel HybridBERT-LSTM architecture that integrates BERT’s contextualized embeddings with LSTM’s sequential processing capabilities to enhance sentiment classification performance in dialogue scenarios. Our framework employs a dual-pooling mechanism to capture local semantic features and global discourse dependencies, addressing limitations of conventional approaches. Comprehensive evaluation on IMDb benchmark and real-world dialogue datasets demonstrates that HybridBERT-LSTM consistently improves over standalone models (LSTM, BERT, CNN, SVM) across accuracy, precision, recall, and F1-score metrics. The architecture effectively exploits pre-trained contextual representations through bidirectional LSTM layers for temporal discourse modeling. We introduce WordContextGraphExplainer, a graph-theoretic interpretability framework addressing conventional explanation method limitations. Unlike LIME’s linear additivity assumptions treating features independently, our approach utilizes perturbation-based analysis to model non-linear semantic interactions. The framework generates semantic interaction graphs with nodes representing word contributions and edges encoding inter-word dependencies, visualizing contextual sentiment propagation patterns. Empirical analysis reveals LIME’s inadequacies in capturing temporal discourse dependencies and collaborative semantic interactions crucial for dialogue sentiment understanding. WordContextGraphExplainer explicitly models semantic interdependencies, negation scope, and temporal flow across conversational turns, enabling comprehensive understanding of both word-level contributions and contextual interaction influences on decision-making processes. This integrated framework establishes a new paradigm for interpretable dialogue sentiment analysis, advancing trustworthy AI through high-performance classification coupled with comprehensive explainability.

由于多回合对话中复杂的语境语义和时间依赖性，自然语言处理中的会话情感分析面临着巨大的挑战。我们提出了一种新的HybridBERT-LSTM架构，它将BERT的上下文嵌入与LSTM的顺序处理能力相结合，以增强对话场景中的情感分类性能。我们的框架采用双池机制来捕获局部语义特征和全局话语依赖，解决了传统方法的局限性。对IMDb基准和真实世界对话数据集的综合评估表明，HybridBERT-LSTM在准确性、精密度、召回率和f1分数指标上持续优于独立模型（LSTM、BERT、CNN、SVM）。该体系结构通过双向LSTM层有效地利用预训练的上下文表示进行时间话语建模。我们介绍了WordContextGraphExplainer，这是一个图论可解释性框架，解决了传统解释方法的局限性。与LIME的线性可加性假设独立处理特征不同，我们的方法利用基于微扰的分析来模拟非线性语义交互。该框架生成语义交互图，节点表示词贡献，边编码词间依赖关系，可视化上下文情感传播模式。实证分析表明，LIME在捕捉对对话情感理解至关重要的时间话语依赖和协作语义交互方面存在不足。WordContextGraphExplainer明确地对语义相互依赖、否定范围和跨会话回合的时间流进行建模，从而能够全面理解单词级别的贡献和上下文交互对决策过程的影响。这个集成框架为可解释的对话情感分析建立了一个新的范式，通过高性能分类和全面的可解释性来推进可信赖的人工智能。

{"title":"Graph-based interpretable dialogue sentiment analysis: A HybridBERT-LSTM framework with semantic interaction explainer","authors":"Ercan Atagün , Günay Temür , Serdar Bi̇roğul","doi":"10.1016/j.csi.2025.104086","DOIUrl":"10.1016/j.csi.2025.104086","url":null,"abstract":"<div><div>Conversational sentiment analysis in natural language processing faces substantial challenges due to intricate contextual semantics and temporal dependencies within multi-turn dialogues. We present a novel HybridBERT-LSTM architecture that integrates BERT’s contextualized embeddings with LSTM’s sequential processing capabilities to enhance sentiment classification performance in dialogue scenarios. Our framework employs a dual-pooling mechanism to capture local semantic features and global discourse dependencies, addressing limitations of conventional approaches. Comprehensive evaluation on IMDb benchmark and real-world dialogue datasets demonstrates that HybridBERT-LSTM consistently improves over standalone models (LSTM, BERT, CNN, SVM) across accuracy, precision, recall, and F1-score metrics. The architecture effectively exploits pre-trained contextual representations through bidirectional LSTM layers for temporal discourse modeling. We introduce WordContextGraphExplainer, a graph-theoretic interpretability framework addressing conventional explanation method limitations. Unlike LIME’s linear additivity assumptions treating features independently, our approach utilizes perturbation-based analysis to model non-linear semantic interactions. The framework generates semantic interaction graphs with nodes representing word contributions and edges encoding inter-word dependencies, visualizing contextual sentiment propagation patterns. Empirical analysis reveals LIME’s inadequacies in capturing temporal discourse dependencies and collaborative semantic interactions crucial for dialogue sentiment understanding. WordContextGraphExplainer explicitly models semantic interdependencies, negation scope, and temporal flow across conversational turns, enabling comprehensive understanding of both word-level contributions and contextual interaction influences on decision-making processes. This integrated framework establishes a new paradigm for interpretable dialogue sentiment analysis, advancing trustworthy AI through high-performance classification coupled with comprehensive explainability.</div></div>","PeriodicalId":50635,"journal":{"name":"Computer Standards & Interfaces","volume":"97 ","pages":"Article 104086"},"PeriodicalIF":3.1,"publicationDate":"2025-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145625273","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Fluent: Round-efficient secure aggregation for private federated learning Fluent：用于私有联邦学习的高效安全聚合

IF 3.1 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computer Standards & Interfaces

Pub Date : 2025-11-12 DOI: 10.1016/j.csi.2025.104096

Xincheng Li , Jianting Ning , Geong Sen Poh , Xinchun Yin , Leo Yu Zhang , Tianwei Zhang

Federated learning (FL) facilitates collaborative training of machine learning models among a large number of clients while safeguarding the privacy of their local datasets. However, FL remains susceptible to vulnerabilities such as privacy inference and inversion attacks. Single-server secure aggregation schemes were proposed to address these threats. Nonetheless, they encounter practical constraints due to their round and communication complexities. This work introduces Fluent, a round-and communication-efficient secure aggregation scheme for private FL. Fluent offers improvements compared to state-of-the-art solutions, i.e., BBGLR (Bell et al., 2020) and Flamingo (Ma et al., 2023): (1) it eliminates the need for iterative handshakes and secret sharing operations by efficiently reusing the shares across multiple training iterations, while maintaining an equivalent level of security assurance as BBGLR and Flamingo; (2) it accomplishes both the consistency check and weight unmasking in one round, thereby reducing the communication complexity. With these constructions, Fluent achieves the fewest communication rounds (i.e., two in the collection phase) in the malicious server setting, in contrast to at least three rounds in BBGLR and Flamingo. This significantly reduces the latency for geographically distributed clients; (3) In addition, a dynamic variant with a participant selection algorithm and multilevel (hierarchical) secret sharing scheme was introduced. It facilitates dynamic client joining, hence enhancing the flexibility and scalability of Fluent. We implemented Fluent and compared it with BBGLR and Flamingo. Experimental results based on the EMNIST dataset show that Fluent achieves significant enhancements in training efficiency. Specifically, Fluent achieves improvements of 6.8

\times

and 2.4

\times

in training time against BBGLR and Flamingo respectively, without compromising accuracy.

联邦学习（FL）促进了大量客户端之间机器学习模型的协作训练，同时保护了本地数据集的隐私。然而，FL仍然容易受到隐私推断和反转攻击等漏洞的影响。针对这些威胁，提出了单服务器安全聚合方案。然而，由于它们的循环和通信复杂性，它们遇到了实际的限制。本工作介绍了Fluent，一种用于私有FL的圆形和通信高效的安全聚合方案。与最先进的解决方案(即BBGLR （Bell等人，2020）和Flamingo （Ma等人，2023）相比，Fluent提供了改进：(1)通过在多个训练迭代中有效地重用共享，消除了迭代握手和秘密共享操作的需要，同时保持了与BBGLR和Flamingo相同的安全保证水平；(2)在一轮内完成一致性检查和权值揭开，降低了通信复杂度。通过这些结构，Fluent在恶意服务器设置中实现了最少的通信轮数（即在收集阶段实现了两轮），而在BBGLR和Flamingo中至少实现了三轮。这大大减少了地理上分布的客户端的延迟；(3)此外，引入了一种具有参与者选择算法和多级（分层）秘密共享方案的动态变体。它促进了动态客户端加入，从而增强了Fluent的灵活性和可扩展性。我们实现了Fluent，并将其与BBGLR和Flamingo进行了比较。基于EMNIST数据集的实验结果表明，Fluent在训练效率上有显著提高。具体来说，在不影响准确率的前提下，Fluent对BBGLR和Flamingo的训练时间分别提高了6.8倍和2.4倍。

{"title":"Fluent: Round-efficient secure aggregation for private federated learning","authors":"Xincheng Li , Jianting Ning , Geong Sen Poh , Xinchun Yin , Leo Yu Zhang , Tianwei Zhang","doi":"10.1016/j.csi.2025.104096","DOIUrl":"10.1016/j.csi.2025.104096","url":null,"abstract":"<div><div>Federated learning (FL) facilitates collaborative training of machine learning models among a large number of clients while safeguarding the privacy of their local datasets. However, FL remains susceptible to vulnerabilities such as privacy inference and inversion attacks. Single-server secure aggregation schemes were proposed to address these threats. Nonetheless, they encounter practical constraints due to their round and communication complexities. This work introduces Fluent, a round-and communication-efficient secure aggregation scheme for private FL. Fluent offers improvements compared to state-of-the-art solutions, i.e., BBGLR (Bell et al., 2020) and Flamingo (Ma et al., 2023): (1) it eliminates the need for iterative handshakes and secret sharing operations by efficiently reusing the shares across multiple training iterations, while maintaining an equivalent level of security assurance as BBGLR and Flamingo; (2) it accomplishes both the consistency check and weight unmasking in one round, thereby reducing the communication complexity. With these constructions, Fluent achieves the fewest communication rounds (i.e., two in the collection phase) in the malicious server setting, in contrast to at least three rounds in BBGLR and Flamingo. This significantly reduces the latency for geographically distributed clients; (3) In addition, a dynamic variant with a participant selection algorithm and multilevel (hierarchical) secret sharing scheme was introduced. It facilitates dynamic client joining, hence enhancing the flexibility and scalability of Fluent. We implemented Fluent and compared it with BBGLR and Flamingo. Experimental results based on the EMNIST dataset show that Fluent achieves significant enhancements in training efficiency. Specifically, Fluent achieves improvements of 6.8<span><math><mo>×</mo></math></span> and 2.4<span><math><mo>×</mo></math></span> in training time against BBGLR and Flamingo respectively, without compromising accuracy.</div></div>","PeriodicalId":50635,"journal":{"name":"Computer Standards & Interfaces","volume":"96 ","pages":"Article 104096"},"PeriodicalIF":3.1,"publicationDate":"2025-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145578669","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Enhanced YOLOv8 model for accurate and real-time remote sensing target detection 增强的YOLOv8模型用于精确和实时的遥感目标检测

IF 3.1 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computer Standards & Interfaces

Pub Date : 2025-11-12 DOI: 10.1016/j.csi.2025.104093

Israr Ahmad , Fengjun Shang , Muhammad Faheem

Current remote sensing image object detection algorithms often struggle with false positives, missed targets, and suboptimal accuracy. To address these issues, we propose an improved YOLOv8 network (PIYN) solution achieved through targeted modifications to the YOLOv8 architecture. The backbone of YOLOv8 utilizes a Cross-Stage Partial (CSP) structure that includes two convolutions, called a faster C2f module. Firstly, we infuse the C2f module integrating an Efficient Multi-Scale Attention (EMA) mechanism, which enhances the module's ability to process information across various scales. Secondly, we introduce a Compact Path Aggregation Network (Compact-PAN) structure within the neck of the network, which reduces the computational complexity of the model. Finally, replacing the Complete Intersection over Union (CIoU) loss function with the Weighted Intersection over Union (WIoU) loss refines the model's detection accuracy. Additionally, we applied K-fold cross-validation on the dataset to mitigate overfitting. Experiments using the extensive Dataset for Object Detection in Aerial images (DOTA) and the Dataset for Object Recognition in Optical Remote Sensing Imagery (DIOR) reveal PIYN's effectiveness: there is a 2.43 % and 2.56 % increase in Mean Average Precision (mAP) over YOLOv8, respectively, alongside a 4.49 % reduction in GFLOPs. These results demonstrate PIYN's capability to enhance accuracy while maintaining efficiency and solidify its progressive and practical impact, particularly for smart city applications.

当前的遥感图像目标检测算法经常存在误报、目标缺失和精度不理想等问题。为了解决这些问题，我们提出了一个改进的YOLOv8网络（PIYN）解决方案，通过对YOLOv8架构进行有针对性的修改来实现。YOLOv8的主干采用了包括两个卷积的跨阶段部分（CSP）结构，称为更快的C2f模块。首先，我们在C2f模块中引入了高效多尺度注意（EMA）机制，增强了模块跨尺度处理信息的能力。其次，在网络颈部引入紧凑路径聚合网络（Compact- pan）结构，降低了模型的计算复杂度；最后，用加权交联（加权交联）损失代替完全交联（CIoU）损失函数，提高了模型的检测精度。此外，我们在数据集上应用K-fold交叉验证来减轻过拟合。使用航空图像中物体检测的广泛数据集（DOTA）和光学遥感图像中物体识别的数据集（DIOR）进行的实验显示了PIYN的有效性：与YOLOv8相比，平均精度（mAP）分别提高了2.43%和2.56%，GFLOPs降低了4.49%。这些结果表明，PIYN能够在保持效率的同时提高准确性，并巩固其进步和实际影响，特别是在智慧城市应用中。

{"title":"Enhanced YOLOv8 model for accurate and real-time remote sensing target detection","authors":"Israr Ahmad , Fengjun Shang , Muhammad Faheem","doi":"10.1016/j.csi.2025.104093","DOIUrl":"10.1016/j.csi.2025.104093","url":null,"abstract":"<div><div>Current remote sensing image object detection algorithms often struggle with false positives, missed targets, and suboptimal accuracy. To address these issues, we propose an improved YOLOv8 network (PIYN) solution achieved through targeted modifications to the YOLOv8 architecture. The backbone of YOLOv8 utilizes a Cross-Stage Partial (CSP) structure that includes two convolutions, called a faster C2f module. Firstly, we infuse the C2f module integrating an Efficient Multi-Scale Attention (EMA) mechanism, which enhances the module's ability to process information across various scales. Secondly, we introduce a Compact Path Aggregation Network (Compact-PAN) structure within the neck of the network, which reduces the computational complexity of the model. Finally, replacing the Complete Intersection over Union (CIoU) loss function with the Weighted Intersection over Union (WIoU) loss refines the model's detection accuracy. Additionally, we applied K-fold cross-validation on the dataset to mitigate overfitting. Experiments using the extensive Dataset for Object Detection in Aerial images (DOTA) and the Dataset for Object Recognition in Optical Remote Sensing Imagery (DIOR) reveal PIYN's effectiveness: there is a 2.43 % and 2.56 % increase in Mean Average Precision (mAP) over YOLOv8, respectively, alongside a 4.49 % reduction in GFLOPs. These results demonstrate PIYN's capability to enhance accuracy while maintaining efficiency and solidify its progressive and practical impact, particularly for smart city applications.</div></div>","PeriodicalId":50635,"journal":{"name":"Computer Standards & Interfaces","volume":"96 ","pages":"Article 104093"},"PeriodicalIF":3.1,"publicationDate":"2025-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145578670","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

BERE: Ensuring immutability and transparency with blockchain in real estate appraisal industry 这里：用区块链保证房地产估价行业的不变性和透明度

IF 3.1 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computer Standards & Interfaces

Pub Date : 2025-11-12 DOI: 10.1016/j.csi.2025.104095

Muruvvet Kalkan , Murat Karakus

Real estate appraisal plays a pivotal role in various financial transactions and regulatory processes, necessitating the assurance of unalterable and transparent Real Estate Appraisal rePorts (REA_Ps). Authorized appraisal experts employ diverse methodologies to ascertain the market value of real estate properties, contributing to critical decisions in property transactions, loans, taxation, insurance, financial reporting, and lease calculations. The standardized format and jurisdiction-compliant methodologies used in official REA_Ps emphasize the need for their preservation in an unaltered state. Despite the importance of official REA_Ps, the potential for unauthorized alterations poses a significant challenge when stored in centralized databases. With its inherent tamper-proof nature, Blockchain (BC) technology emerges as an ideal solution to safeguard integrity of the REA_Ps. This study proposes a Blockchain-Enhanced Real Estate appraisal management framework, BERE, to ensure the immutability of REA_Ps and enhance appraisal process transparency. The proposed BERE framework fosters trust among stakeholders in the real estate appraisal sector. By securely storing REA_Ps authored by on-field appraisal experts on the BC, the method prevents unauthorized modifications and ensures the confidentiality of private data. Simulations employ a variety of measures, such as BC size, transaction verification time, CPU usage, and message overhead on transaction generation, to assess system performance. The results indicate that the BERE framework successfully maintains the accuracy and reliability of REA_Ps. To the best of our knowledge, this is the first comprehensive study showcasing a practical application of BC technology to demonstrate its ability to transform the appraisal process and inspire trust in the accuracy of REA_Ps.

房地产评估在各种金融交易和监管过程中发挥着关键作用，因此需要确保房地产评估报告（REA_Ps）的不可更改和透明。授权评估专家采用多种方法来确定房地产的市场价值，为房地产交易、贷款、税收、保险、财务报告和租赁计算的关键决策做出贡献。官方rea_p中使用的标准化格式和符合管辖权的方法强调需要将它们保持在不变的状态。尽管官方rea_p很重要，但是当存储在集中式数据库中时，未经授权的更改的可能性构成了重大挑战。由于其固有的防篡改特性，区块链（BC）技术成为保护rea_p完整性的理想解决方案。本研究提出了一个区块链增强的房地产评估管理框架，即here，以确保rea_p的不变性，并提高评估过程的透明度。提议的BERE框架促进了房地产评估部门利益相关者之间的信任。通过将现场评估专家撰写的rea_p安全地存储在BC上，该方法可以防止未经授权的修改，并确保私人数据的机密性。模拟使用各种度量，例如BC大小、事务验证时间、CPU使用情况和事务生成时的消息开销，来评估系统性能。结果表明，该框架成功地保持了REA_Ps的准确性和可靠性。据我们所知，这是第一个全面的研究，展示了BC技术的实际应用，展示了它改变评估过程的能力，并激发了对rea_p准确性的信任。

{"title":"BERE: Ensuring immutability and transparency with blockchain in real estate appraisal industry","authors":"Muruvvet Kalkan , Murat Karakus","doi":"10.1016/j.csi.2025.104095","DOIUrl":"10.1016/j.csi.2025.104095","url":null,"abstract":"<div><div>Real estate appraisal plays a pivotal role in various financial transactions and regulatory processes, necessitating the assurance of unalterable and transparent Real Estate Appraisal rePorts (REA_Ps). Authorized appraisal experts employ diverse methodologies to ascertain the market value of real estate properties, contributing to critical decisions in property transactions, loans, taxation, insurance, financial reporting, and lease calculations. The standardized format and jurisdiction-compliant methodologies used in official REA_Ps emphasize the need for their preservation in an unaltered state. Despite the importance of official REA_Ps, the potential for unauthorized alterations poses a significant challenge when stored in centralized databases. With its inherent tamper-proof nature, Blockchain (BC) technology emerges as an ideal solution to safeguard integrity of the REA_Ps. This study proposes a Blockchain-Enhanced Real Estate appraisal management framework, <em>BERE</em>, to ensure the immutability of REA_Ps and enhance appraisal process transparency. The proposed <em>BERE</em> framework fosters trust among stakeholders in the real estate appraisal sector. By securely storing REA_Ps authored by on-field appraisal experts on the BC, the method prevents unauthorized modifications and ensures the confidentiality of private data. Simulations employ a variety of measures, such as BC size, transaction verification time, CPU usage, and message overhead on transaction generation, to assess system performance. The results indicate that the <em>BERE</em> framework successfully maintains the accuracy and reliability of REA_Ps. To the best of our knowledge, this is the first comprehensive study showcasing a practical application of BC technology to demonstrate its ability to transform the appraisal process and inspire trust in the accuracy of REA_Ps.</div></div>","PeriodicalId":50635,"journal":{"name":"Computer Standards & Interfaces","volume":"96 ","pages":"Article 104095"},"PeriodicalIF":3.1,"publicationDate":"2025-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145578672","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

How AI agents transform reflective practices: A three-semester comparative study in socially shared regulation of learning 人工智能代理如何转变反思实践：社会共享学习规则的三个学期比较研究

IF 3.1 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computer Standards & Interfaces

Pub Date : 2025-11-11 DOI: 10.1016/j.csi.2025.104094

Yumin Zheng , Fengjiao Tu , Fengfang Shu , Chaowang Shang , Lulu Chen , Jiang Meng

High-quality reflection has been a challenging barrier in the socially shared regulation of learning (SSRL). Especially with the emergence of generative artificial intelligence (GAI), traditional methods such as reflection reports may increase the students' risk of superficial reflection. This study uses an artificial intelligence agent (AI agent) to design a reflection assistant, which aims to enhance students' reflection ability through continuous questioning and real-time, content-specific feedback based on their written reflections. Through a comparative experiment conducted over three semesters, this study demonstrates the different impacts of three reflection methods, reflection reports, reflection short-answer questions, and AI agents, on the quality of university students' reflections. The results indicate that there is a significant difference in the quality of reflection among the three reflection methods. Students using AI agents show the highest levels of reflection, characterized primarily by connective reflection and critical reflection. Epistemic network analysis further reveals that the AI agent reflection method is more effective in improving the reflection quality of low-performance teams than that of high-performance teams. This expands AI agents' use in SSRL reflection, introduces new methods for the GAI era, and provides practical experience and reflection intervention strategies for teachers and instructional designers in SSRL.

高质量的反思一直是社会共享学习调节（SSRL）的一个具有挑战性的障碍。特别是随着生成式人工智能（GAI）的出现，传统的反思报告等方法可能会增加学生肤浅反思的风险。本研究使用人工智能代理（AI agent）设计了一个反思助手，旨在通过持续的提问和基于学生书面反思的实时、具体内容的反馈来增强学生的反思能力。本研究通过三个学期的对比实验，论证了反思报告、反思简答题和人工智能代理三种反思方式对大学生反思质量的不同影响。结果表明，三种反射方法的反射质量存在显著差异。使用人工智能代理的学生表现出最高水平的反思，主要表现为关联反思和批判性反思。认知网络分析进一步揭示了人工智能主体反思方法在提高低绩效团队反思质量方面比高绩效团队反思质量更有效。这拓展了AI agent在SSRL反思中的应用，为GAI时代引入了新的方法，为SSRL教师和教学设计师提供了实践经验和反思干预策略。

{"title":"How AI agents transform reflective practices: A three-semester comparative study in socially shared regulation of learning","authors":"Yumin Zheng , Fengjiao Tu , Fengfang Shu , Chaowang Shang , Lulu Chen , Jiang Meng","doi":"10.1016/j.csi.2025.104094","DOIUrl":"10.1016/j.csi.2025.104094","url":null,"abstract":"<div><div>High-quality reflection has been a challenging barrier in the socially shared regulation of learning (SSRL). Especially with the emergence of generative artificial intelligence (GAI), traditional methods such as reflection reports may increase the students' risk of superficial reflection. This study uses an artificial intelligence agent (AI agent) to design a reflection assistant, which aims to enhance students' reflection ability through continuous questioning and real-time, content-specific feedback based on their written reflections. Through a comparative experiment conducted over three semesters, this study demonstrates the different impacts of three reflection methods, reflection reports, reflection short-answer questions, and AI agents, on the quality of university students' reflections. The results indicate that there is a significant difference in the quality of reflection among the three reflection methods. Students using AI agents show the highest levels of reflection, characterized primarily by connective reflection and critical reflection. Epistemic network analysis further reveals that the AI agent reflection method is more effective in improving the reflection quality of low-performance teams than that of high-performance teams. This expands AI agents' use in SSRL reflection, introduces new methods for the GAI era, and provides practical experience and reflection intervention strategies for teachers and instructional designers in SSRL.</div></div>","PeriodicalId":50635,"journal":{"name":"Computer Standards & Interfaces","volume":"97 ","pages":"Article 104094"},"PeriodicalIF":3.1,"publicationDate":"2025-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145580387","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0