检索结果-内蒙古大学图书馆

RanViz: Ransomware Visualization and Classification Based on Time-Series Categorical Representation of API Calls

IEEE ACCESS 2025年 13卷 56237-56254页

作者： Mokoma, Vhuhwavho Singh, Avinash Univ Pretoria Dept Comp Sci ZA-0002 Pretoria South Africa

Ransomware continues to pose a significant threat to individuals and organizations worldwide, causing disruptions, financial losses, and reputational damage. As ransomware attacks grow in sophistication, understanding their behaviour through effective analysis has become increasingly critical for mitigation and prevention. However, ransomware analysis presents several challenges. First, the sheer volume of application programming Interface (API) call data generated by ransomware during execution can overwhelm traditional analysis methods. Second, the temporal and categorical nature of this data makes identifying meaningful patterns complex. Third, the integration of machine learning (ML) models, which are essential for accurate classification, is hindered by the difficulty of modelling intricate API call behaviours. Without effective tools to address these issues, analysts risk missing critical behavioural indicators. To overcome these challenges, the proposed Ransomware Visualization (RanViz) system was developed to provide a comprehensive visual analytics and classification platform designed to enhance ransomware analysis. RanViz employs advanced visualization techniques to represent categorical API call time-series data, enabling analysts to intuitively understand ransomware behaviours that might otherwise remain obscured. The system incorporates ML models based on API call frequency, temporal interval, and sequence to classify unknown samples as either benign or ransomware. The models collectively achieve an accuracy of over 95% in detecting ransomware. By providing a unified platform that combines powerful visualization tools with high-performing ML models, RanViz simplifies ransomware analysis and offers a robust framework for accurate classification. This makes it an invaluable tool for digital forensics and cybersecurity professionals tasked with addressing the ever-evolving ransomware threat.

关键词： Ransomware Malware Data visualization application programming interfaces Encryption Visual analytics Organizations Computer security Complexity theory Ransomware analysis API calls machine learning ransomware time series visualization visualization

来源：评论

学校读者我要写书评

暂无评论

Web API Recommendation via Exploring Textual and Structural Semantics With Contrastive Learning and Joint Training

引用

IEEE TRANSACTIONS ON NETWORK AND SERVICE MANAGEMENT 2025年第2期22卷 1558-1568页

作者： Kang, Guosheng Ren, Hongshuai Chen, Wanjun Liu, Jianxun Cao, Buqing Xu, Yu Hunan Univ Sci & Technol Sch Comp Sci & Engn Xiangtan 411199 Peoples R China Yunnan Univ Finance & Econ Yunnan Key Lab Serv Comp Kunming 650503 Peoples R China Hunan Univ Sci & Technol Hunan Prov Key Lab Serv Comp & Novel Software Tech Xiangtan 411199 Peoples R China Univ Cambridge Dept Publ Hlth & Primary Care Cambridge Baker Syst Genom Initiat Cambridge CB2 1TN England Univ Cambridge British Heart Fdn Cardiovasc Epidemiol Unit Dept Publ Hlth & Primary Care Cambridge CB2 1TN England

With the advancement of service computing technology, software developers tend to consume a variety of Web APIs (application programming interfaces, also named Web services) from Web API markets to create feature-rich Mashup applications to save time and cost. Under such a background, the ever-increasing number of Web APIs makes the service discovery become a challenge. Thus, Web API recommendation becomes an effective means for service discovery. However, the existing approaches to Web API recommendation still have limitations in extracting rich semantics sufficiently from functional description documents and service networks, resulting in a limited recommendation performance. To further improve the recommendation performance, this paper proposes an effective Web API recommendation approach via exploring textual and structural semantics with contrastive learning and joint training, named CLJT. On one side, discriminative feature representations from textual and structural semantics could be derived by contrastive learning with information correlation across views. On the other side, the derived representations could be applicable to Web API recommendation by joint training of the representation tasks and the recommendation task. Extensive experiments are conducted over a real-world dataset crawled from ***. The experimental results demonstrate the superiority of the proposed approach compared to the baseline methods.

关键词： Semantics Mashups application programming interfaces Training Encoding Contrastive learning Bidirectional control Service computing Natural language processing Vectors Mashup creation Web API recommendation textual semantic structural semantic contrastive learning joint training

来源：评论

学校读者我要写书评

暂无评论

Artificial intelligence boundary resources: a relational view on leveraging "AI-as-a-Service"

引用

EUROPEAN JOURNAL OF INFORMATION SYSTEMS 2025年

作者： Hanelt, Andre Firk, Sebastian Zapadka, Patryk Oehmichen, Jana Univ Kassel Chair Digital Transformat Management Kassel Germany Univ Groningen Dept Accounting & Auditing Groningen Netherlands Johannes Gutenberg Univ Mainz Human Resources & Management Studies Mainz Germany

In response to the increasing relevance of artificial intelligence (AI) for competitive advantage, firms can utilize the capabilities of AI-savvy partners via boundary resources. Due to AI facets, such as autonomy, inscrutability, and learning, this strategy holds substantial promises for productivity and innovation, yet is not free of risks, given the strategic importance of AI and potential dependencies in this key area. As the antecedents and consequences of this strategy are unclear, we build a theoretical framework about these aspects based on the relational view of competitive advantage. We test our hypotheses on a longitudinal dataset and find that internal AI knowledge and the presence of a chief information officer in the top management team are associated with selecting AI boundary resources for process improvements. High external market pressure exerted by digital ventures and the AI sophistication of industry peers are positively associated with selecting AI boundary resources for product improvements. We further find that the use of AI boundary resources for process improvements is positively associated with operational efficiency and the use of AI boundary resources for product improvements with increasing sales. We derive implications for IS research on managing AI and boundary resources, as well as managerial practice.

关键词： Boundary resources application programming interfaces artificial intelligence CIO relational view

来源：评论

学校读者我要写书评

暂无评论

LLM-Driven, Self-Improving Framework for Security Test Automation: Leveraging Karate DSL for Augmented API Resilience

引用

IEEE ACCESS 2025年 13卷 56861-56886页

作者： Pasca, Emil Marian Delinschi, Daniela Erdei, Rudolf Matei, Oliviu Tech Univ Cluj Napoca North Univ Ctr Baia Mare Dept Elect Elect & Comp Engn Baia Mare 430122 Romania HOLISUN SRL Dept Res & Dev Baia Mare 430397 Romania

Modern software architectures heavily rely on APIs, yet face significant security challenges, particularly with Broken Object Level Authorization (BOLA) vulnerabilities, which remain the most critical API security risk according to OWASP. This paper introduces Karate-BOLA-Guard, an innovative framework leveraging Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG) techniques to automate security-focused test case generation for APIs. Our approach integrates vector databases for context retrieval, multiple LLM models for test generation, and observability tools for process monitoring. Initial experiments were carried out on three deliberately vulnerable APIs (VAmPI, Crapi, and OWASP Juice Shop), with subsequent validation on fifteen additional production APIs spanning diverse domains including social media, version control systems, financial services, and transportation services. Our evaluation metrics show Llama 3 8B achieving consistent performance (Accuracy: 3.1-3.4, Interoperability: 3.7-4.3) with an average processing time of 143.76 seconds on GPU. Performance analysis revealed significant GPU acceleration benefits, with 20-25x improvement over CPU processing times. Smaller models demonstrated efficient processing, with Phi-3 Mini averaging 69.58 seconds and Mistral 72.14 seconds, while maintaining acceptable accuracy scores. Token utilization patterns showed Llama 3 8B using an average of 36,591 tokens per session, compared to Mistral's 25,225 and Phi-3 Mini's 31,007. Our framework's effectiveness varied across APIs, with notably strong performance in complex platforms (Instagram: A = 4.3, I = 4.4) while maintaining consistent functionality in simpler implementations (VAmPI: A = 3.6, I = 4.3). The iterative refinement process, evaluated through comprehensive metrics including Accuracy (A), Complexity (C), and Interoperability (I), represents a significant advancement in automated API security testing, offering an efficient, accurate, and adapt

关键词： Security Testing Retrieval augmented generation Test pattern generators application programming interfaces Accuracy Software testing Automation Systematics Computer architecture API security automation testing tools cybersecurity restful API software testing

来源：评论

学校读者我要写书评

暂无评论

DataPoll: A Tool Facilitating Big Data Research in Social Sciences

引用

IEEE TRANSACTIONS ON COMPUTATIONAL SOCIAL SYSTEMS 2025年第2期12卷 511-524页

作者： Charalampous, Antonis Djouvas, Constantinos Christodoulou, Christos Cyprus Univ Technol Dept Commun & Internet Studies CY-3036 Limassol Cyprus

The computational analysis of big data has revolutionized social science research, offering unprecedented insights into societal behaviors and trends through digital data from online sources. However, existing tools often face limitations such as technical complexity, single-source dependency, and a narrow range of analytical capabilities, hindering accessibility and effectiveness. This article introduces DataPoll, an end-to-end big data analysis platform designed to democratize computational social science research. DataPoll simplifies data collection, analysis, and visualization, making advanced analytics accessible to researchers of diverse expertise. It supports multisource data integration, innovative analytical features, and interactive dashboards for exploratory and comparative analyses. By fostering collaboration and enabling the integration of new data sources and analysis methods, DataPoll represents a significant advancement in the field. A comprehensive case study on the Ukrainian--Russian conflict demonstrates its capabilities, showcasing how DataPoll can yield actionable insights into complex social phenomena. This tool empowers researchers to harness the potential of big data for impactful and inclusive research.

关键词： Big Data Data models Social sciences Pipelines Data mining application programming interfaces Data visualization Crawlers Social networking (online) Machine learning Big data in social sciences big data systems computational methods computational social science online platforms social computing

来源：评论

学校读者我要写书评

暂无评论

Facial recognition by cloud-based APIs following surgically assisted rapid maxillary expansion

引用

Journal of Orofacial Orthopedics 2025年第2期86卷 98-107页

作者： Buyukcavus, Muhammed Hilmi Aydogan Akgun, Filiz Solak, Serdar Ucar, Mustafa Hikmet Bilgehan Fındık, Yavuz Baykul, Timucin Faculty of Dentistry Department of Orthodontics Antalya Bilim University Antalya Turkey Faculty of Dentistry Department of Orthodontics Burdur Mehmet Akif Ersoy University Burdur Turkey Faculty of Technology Department of Information Systems Engineering Kocaeli University Kocaeli Turkey Faculty of Dentistry Department of Department of Oral and Maxillofacial Surgery Süleyman Demirel University Isparta Turkey

Introduction: This study aimed to investigate whether the facial soft tissue changes of individuals who had undergone surgically assisted rapid maxillary expansion (SARME) would be detected by three different well-known facial biometric recognition applications. Methods: To calculate similarity scores, the pre- and postsurgical photographs of 22 patients who had undergone SARME treatment were examined using three prominent cloud computing-based facial recognition application programming interfaces (APIs): AWS Rekognition (Amazon Web Services, Seattle, WA, USA), Microsoft Azure Cognitive (Microsoft, Redmond, WA, USA), and Face++ (Megvii, Beijing, China). The pre- and post-SARME photographs of the patients (relaxed, smiling, profile, and semiprofile) were used to calculate similarity scores using the APIs. Friedman’s two-way analysis of variance and the Wilcoxon signed-rank test were used to compare the similarity scores obtained from the photographs of the different aspects of the face before and after surgery using the different programs. The relationship between measurements on lateral and posteroanterior cephalograms and the similarity scores was evaluated using the Spearman rank correlation. Results: The similarity scores were found to be lower with the Face++ program. When looking at the photo types, it was observed that the similarity scores were higher in the smiling photos. A statistically significant difference in the similarity scores (P < 0.05) was found between the relaxed and smiling photographs using the different programs. The correlation between the cephalometric and posteroanterior measurements and the similarity scores was not significant (P > 0.05). Conclusion: SARME treatment caused a significant change in the similarity scores calculated with the help of three different facial recognition programs. The highest similarity scores were found in the smiling photographs, whereas the lowest scores were found in the profile photographs. © 2023, The Auth

关键词： Amazon Web Services application programming interfaces Artificial intelligence Azure Face++

来源：评论

学校读者我要写书评

暂无评论

Label Space-Induced Pseudo Label Refinement for Multi-Source Black-Box Domain Adaptation

引用

IEEE Transactions on Image Processing 2025年 34卷 3181-3193页

作者： Yoo, Chaehwa Liu, Xiaofeng Xing, Fangxu Woo, Jonghye Kang, Je-Won Chungbuk National University School of Electrical Engineering Cheongju Korea Republic of Massachusetts General Hospital Harvard Medical School Gordon Center for Medical Imaging Dept. of Radiology BostonMA United States Yale University New Haven United States Ewha W. University Department of Electronic and Electrical Engineering Graduate Program in Smart Factory Seoul Korea Republic of

Conventional unsupervised domain adaptation (UDA) requires access to source data and/or source model parameters, prohibiting its practical application in terms of privacy, security, and intellectual property. Recent black-box UDA (BDA) reduces such constraints by defining a pseudo label from a single encapsulated source application programming interface (API) prediction, which allows for self-training of the target model. Nonetheless, existing methods have limited consideration for multi-source settings, in which multiple source domain APIs are available to generate pseudo labels. In this work, we introduce a novel training framework for multi-source BDA (MSBDA), dubbed Label Space-Induced Pseudo Label Refinement (LPR). Specifically, LPR incorporates a Pseudo label Refinery Network (PRN) that learns the relationship among source domains conditioned by the target domain only utilizing source API’s prediction. The target model is adapted by our dual phases PRN. First, a warm-up phase targets to avoid failure due to noisy samples and provide an initial pseudo-label, which is followed by a label refinement phase with domain relationship exploration. We provide theoretical support for the mechanism of the LPR. Experimental results on four benchmark datasets demonstrate that MSBDA using LPR achieves competitive performance compared to state-of-the-art approaches with different DA settings. © 1992-2012 IEEE.

关键词： Adaptation models Labeling Data models Predictive models Training Closed box application programming interfaces Upper bound Sports Smoothing methods

来源：评论

学校读者我要写书评

暂无评论

Honeytrap Resilience: A Study of Malware Redirection Awareness

Honeytrap Resilience: A Study of Malware Redirection Awarene...

引用

International Symposium on Digital Forensic and Security (ISDFS)

作者： Eric Edge James H Jones Kathryn Laskey George Mason University

ISBN: (数字)9798331509934

ISBN: (纸本)9798331509941

This research investigates the ability of self-propagating malware to detect network redirection, a technique that reroutes malicious traffic to a controlled environment for analysis. We hypothesized that malware programs might use round-trip time (RTT) as a signal to detect redirection. To test this, we executed a large dataset of malware samples in a controlled environment and analyzed their behavior, measured by application programming interface (API) call patterns. Our results indicate that the malware samples were not sensitive to network redirection, as there were no significant differences in their behavior between redirected and nonredirected scenarios. These findings suggest that network redirection remains a valuable tool for analyzing malicious traffic, even in the face of potential evasion techniques.

关键词： Systematics Sensitivity Digital forensics Routing Malware Security Faces application programming interfaces Resilience

来源：评论

学校读者我要写书评

暂无评论

Securing QR Codes Infrastructure Using AI to Detect Malicious Activity

Securing QR Codes Infrastructure Using AI to Detect Maliciou...

引用

IEEE Annual Computing and Communication Workshop and Conference (CCWC)

作者： Luis Rivas Varun Kumar Singh Vinayak Khandelwal Lanier Watkins US Army Johns Hopkins University Information Security Institute Johns Hopkins University

ISBN: (数字)9798331507695

ISBN: (纸本)9798331507701

With the advent of QR Codes, mobile devices, and computers, users can exchange information quickly but not se-curely. QR Codes-based attacks are rising rapidly, threatening the trust of this extensively used technology. Whether you use it at a restaurant, conference, or payment station, QR Codes inherently rely on user trust to safeguard transactions. Phishing scams using fake codes have increased by over 50%, while using QR vector malware attacks have increased by over 2400% since 2020 [3]–[5]. These alarming results suggest that integrity protections must be reinstated to restore public trust. This research addresses these challenges by leveraging Artificial intelligence (AI) and third-party services to detect QR Codes-based threats. We propose a Secure QR Codes system employing anomaly detection, risk signaling, and four machine-learning models trained on standard and malicious QR Codes. Using encryption engines, AI-based predictions, application programming Interface (API), proxy services, and AI assessment modules, the integrated solution identified over 96% simulated real-world attacks like phishing, exfiltration, injection, and code execution. Our approach detects QR Codes with modified structures. It provides a practical, non-proprietary, web-deployable solution to predict and mitigate these threats, restoring trust in QR Codes systems across various applications.

关键词： Codes Phishing QR codes Machine learning Vectors Mobile handsets Malware Protection Standards application programming interfaces

来源：评论

学校读者我要写书评

暂无评论

Patterns on Designing API Endpoint Operations 21

Patterns on Designing API Endpoint Operations

引用

Proceedings of the 28th Conference on Pattern Languages of Programs

作者： Apitchaka Singjai Uwe Zdun Olaf Zimmermann Mirko Stocker Cesare Pautasso University of Vienna Software Architecture Research Group Austria OST Eastern Switzerland University of Applied Sciences Switzerland Software Institute Faculty of Informatics USI Lugano Switzerland

ISBN: (纸本)9781941652176

Domain-driven design (DDD) is often applied when implementing microservices or communicating through APIs in distributed systems. APIs expose a published language that provides a view on entire domain models or subsets of such models. Hence, tactical DDD patterns such as Aggregate, Service, and Entity may not only structure API implementations, but also guide API specification work. In our previous work, we described endpoint-level patterns for this context. In this paper, we present three complementary patterns, namely Aggregated Domain Operation on API Endpoint, Event-Based API Endpoint Operation, and CRUD-Based API Operation. These patterns aim to derive API operations from the operations of Domain Services and Entities as well as Domain Events. We also discuss variants of these patterns, such as their combination with the patterns Command Query Responsibility Segregation (CQRS) and Publish/Subscribe. Our pattern mining work is based on a data set from an empirical study of 32 grey literature sources investigating practitioner views on deriving API designs from DDD models.

关键词： application programming interfaces

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：