Article
Computer Science, Information Systems
Xinyu Jiang, Xiangyu Liu, Jiahao Fan, Xinming Ye, Chenyun Dai, Edward A. Clancy, Dario Farina, Wei Chen
Summary: This study proposes a cancelable biometric modality based on high-density surface electromyogram (HD-sEMG) encoded by hand gesture password for user authentication in wireless body area network-based Internet-of-Things applications. The use of automatically generated password-specific channel mask reduces data acquisition and transmission burden in IoT devices. The HD-sEMG biometrics were robust with reduced sampling rate, achieving low equal error rates for both wrong and correct gesture password entries, and providing a cancelable template option for users if needed.
IEEE INTERNET OF THINGS JOURNAL
(2021)
Article
Computer Science, Theory & Methods
Sumaiya Shomaji, Pallabi Ghosh, Fatemeh Ganji, Damon Woodard, Domenic Forte
Summary: A novel biometric framework called HBF was proposed for resource-constrained environments, offering compact storage and fast query processing. This paper explores attack vectors that could compromise the security of HBF and evaluates its security under these well-defined attack vectors through quantitative analyses and experiments, concluding that it is more difficult to attack than traditional Bloom Filters. Additionally, it is found that soft biometric information is also kept private in the HBF-based system.
IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY
(2021)
Review
Computer Science, Information Systems
El-Sayed A. El-Dahshan, Mahmoud M. Bassiouni, Septavera Sharvia, Abdel-Badeeh M. Salem
Summary: This study provides a survey of technologies and methodologies used in PCG biometric systems, covering data acquisition, feature extraction, classification, and evaluation. Out of 157 manuscripts from 2006 to 2020, 35 focused on heart sounds as a biometric, with 11 meeting the inclusion criteria. The results suggest promising prospects for the use of PCG signals, although there are still limitations and future research directions to consider.
COMPUTER SCIENCE REVIEW
(2021)
Article
Computer Science, Information Systems
Shun-Chi Wu, Pei-Lun Hung, A. Lee Swindlehurst
Summary: The article proposes a novel ECG-based biometric recognition scheme to enhance the security of IoT systems. Through subspace oversampling, unique and irreversible templates are created to avoid cross-matching issues and privacy invasion. Experimental results demonstrate the performance and effectiveness of the proposed scheme.
IEEE INTERNET OF THINGS JOURNAL
(2021)
Review
Computer Science, Information Systems
Juan Carlos Bernal-Romero, Juan Manuel Ramirez-Cortes, Jose De Jesus Rangel-Magdaleno, Pilar Gomez-Gil, Hayde Peregrina-Barreto, Israel Cruz-Vega
Summary: This document examines the vulnerabilities of biometric data and proposes countermeasures for the protection and confidentiality of such information. It also proposes a taxonomy and evaluation measures for protection techniques, and highlights the advantages of injective and linear mapping for authentication and identification systems. Additionally, the document mentions commercial products and suggests future directions for cancelable biometric systems in low-cost IoT devices.
Article
Computer Science, Information Systems
Guozhu Zhao, Pinchang Zhang, Yulong Shen, Xiaohong Jiang
Summary: This article proposes a passive authentication framework for continuous user authentication in industrial Internet of Things systems, using behavioral biometrics. Experimental results validate the stability and discriminability of the framework, and various techniques are employed for noise elimination and dimensionality reduction. Extensive experiments are conducted to evaluate the authentication performance and related efficiency issues.
IEEE INTERNET OF THINGS JOURNAL
(2021)
Article
Computer Science, Information Systems
Shahriar Ebrahimi, Siavash Bayat-Sarmadi
Summary: This study proposes a lightweight hardware/software co-design architecture for implementing LPN-based FE, which is resistant to simple side-channel analysis and significantly improves resource requirements compared to previous works.
IEEE INTERNET OF THINGS JOURNAL
(2021)
Review
Computer Science, Information Systems
Riseul Ryu, Soonja Yeom, Soo-Hyung Kim, David Herbert
Summary: Building safeguards against illegitimate access and authentication is crucial for system security. Continuous multimodal biometric authentication systems have been proposed as a reliable solution to address the challenges in existing user authentication schemes. However, there is a lack of critical analysis on current progress in the field, highlighting the need for further research and development in this area.
Article
Computer Science, Information Systems
Jinani Sooriyaarachchi, Suranga Seneviratne, Kanchana Thilakarathna, Albert Y. Zomaya
Summary: MusicID is an authentication solution for smart devices that uses music-induced brainwave patterns as a behavioral biometric modality, achieving high accuracy rates. Experimental results show that data collected from a 4-electrode brainwave headset can achieve over 98% accuracy for user identification and over 97% accuracy for user verification.
IEEE INTERNET OF THINGS JOURNAL
(2021)
Article
Computer Science, Information Systems
Guozhu Zhao, Pinchang Zhang, Yulong Shen, Limei Peng, Xiaohong Jiang
Summary: This article proposes a novel two-dimensional passive authentication framework for Industrial Internet of Things (IIoT) systems, by utilizing the time-varying characteristics of user operation actions and the spatial variation characteristics of channel state information (CSI). The framework includes two classifiers and assigns appropriate weights, enabling continuous and non-intrusive user authentication.
IEEE TRANSACTIONS ON CLOUD COMPUTING
(2023)
Article
Computer Science, Information Systems
Mohammed Abuhamad, Ahmed Abusnaina, Daehun Nyang, David Mohaisen
Summary: The importance of security and user authentication on mobile devices is increasing, with embedded sensors capturing behavioral biometrics for continuous and implicit user authentication.
IEEE INTERNET OF THINGS JOURNAL
(2021)
Article
Multidisciplinary Sciences
Ming-Chin Chuang, Chia-Cheng Yen
Summary: The paper presents a new geometric authentication mechanism that combines geometric characteristics with user biometrics to enhance security in IoT environment.
Article
Automation & Control Systems
Fadi Al-Turjman, B. D. Deebak
Summary: Artificial intelligence-based Internet of Things enables autonomous communication between social networks and IoT, providing a promising solution in modern paradigms. The recent expansion of information-centric networking has introduced a remarkable technique, the public auditing scheme, for IoT-enabled sensor technologies, utilizing cloud-based medical cyber-physical system (M-CPS) for fast computing and reliable data storage.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Computer Science, Information Systems
Meennapa Rukhiran, Sethapong Wong-In, Paniti Netinant
Summary: The traditional human proctoring approach to student identity verification can be unreliable and time-consuming. This study proposes using the internet of things to develop flexible biometric recognition systems and evaluates their effectiveness compared to traditional methods. The results show that unimodal facial and fingerprint biometric systems are suitable for student identity verification, while multimodal and semi-multimodal systems offer higher accuracy with shorter processing times and higher costs.
Article
Computer Science, Artificial Intelligence
Sandeep Gupta, Carsten Maple, Bruno Crispo, Kiran Raja, Artsiom Yautsiukhin, Fabio Martinelli
Summary: This article presents a survey on HCI-based and natural habits-based behavioural biometrics for user recognition in IoT systems. Robust and usable user recognition is crucial for the security of emerging IoT ecosystems. Biometrics provide a solution to the limitations of conventional recognition schemes, and this article reviews the state-of-the-art research on touch-stroke, swipe, touch signature, hand-movements, voice, gait, and footstep behavioural biometrics modalities. The article also explores the security, privacy, and usability evaluations to enhance the design of user recognition schemes for IoT applications.
PATTERN RECOGNITION
(2023)
Article
Computer Science, Information Systems
Zhiyong Yu, Xiangping Zheng, Fangwan Huang, Wenzhong Guo, Lin Sun, Zhiwen Yu
Summary: This study proposes a framework for time series prediction based on sparse representation model (SRM) in smart cities, which can flexibly and effectively handle the challenge of event alert based on time series prediction.
FRONTIERS OF COMPUTER SCIENCE
(2021)
Article
Computer Science, Information Systems
Yixuan Zhang, Jiaqi Liu, Bin Guo, Zhu Wang, Yunji Liang, Zhiwen Yu
Summary: App popularity prediction is a significant task in mobile service development. This paper proposes DeePOP, a popularity prediction model that leverages time-varying hierarchical interactions. By integrating internal factors and time-varying hierarchical interactions, DeePOP achieves higher prediction accuracy compared to existing methods.
IEEE TRANSACTIONS ON MOBILE COMPUTING
(2022)
Editorial Material
Computer Science, Hardware & Architecture
Sagar Samtani, Hsinchun Chen, Murat Kantarcioglu, Bhavani Thuraisingham
IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING
(2022)
Article
Computer Science, Artificial Intelligence
Yasan Ding, Bin Guo, Yan Liu, Yunji Liang, Haocheng Shen, Zhiwen Yu
Summary: The proliferation of fake news on social networks has had devastating impacts on society, the economy, and public security. Existing studies in automatic fake news detection primarily use deep neural networks to learn event-specific features, resulting in limited applicability to upcoming events due to distribution discrepancies. This paper proposes an end-to-end adversarial adaptation network, called MetaDetector, to transfer meta knowledge between different events for fake news detection. Experimental results on real-world datasets show that MetaDetector outperforms state-of-the-art methods, especially in scenarios with significant distribution discrepancies.
ACM TRANSACTIONS ON INTELLIGENT SYSTEMS AND TECHNOLOGY
(2022)
Article
Computer Science, Hardware & Architecture
Yunji Liang, Yuchen Qin, Qi Li, Xiaokai Yan, Luwen Huangfu, Sagar Samtani, Bin Guo, Zhiwen Yu
Summary: With the increasing prevalence of mobile devices, concerns about privacy breaches and data leakage are growing. Previous studies have shown that motion sensors can cause confidential information to be leaked, but the possibility of synthesizing intelligible speech waveforms from low-resolution motion sensors has not yet been thoroughly studied.
IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING
(2023)
Article
Computer Science, Information Systems
Yi Ouyang, Bin Guo, Qianru Wang, Yunji Liang, Zhiwen Yu
Summary: We propose a novel model called dynamic usage graph network (DUGN) to recommend the next app that a user is most likely to use based on their app usage behaviors. By explicitly modeling the complex correlations between apps using a dynamic graph structure, we can learn the dynamics of user interests over time. Experimental results show that our model outperforms existing recommendation methods.
IEEE TRANSACTIONS ON MOBILE COMPUTING
(2023)
Article
Computer Science, Information Systems
Erkang Jing, Yezheng Liu, Yidong Chai, Jianshan Sun, Sagar Samtani, Yuanchun Jiang, Yang Qian
Summary: This paper proposes a deep learning-based speech emotion recognition model that achieves active interpretation of the model results by introducing interpretability and uncorrelation constraints. The model outperforms baselines in experiments and can also learn human perception patterns of speech emotion and provide explanations for recognition results.
INFORMATION PROCESSING & MANAGEMENT
(2023)
Proceedings Paper
Computer Science, Interdisciplinary Applications
Aijia Yuan, Michael Xu, Hongyi Zhu, Sagar Samtani, Edlin Garcia
Summary: This study explores the performance of machine learning models for depression detection using sensors that do not release Personally Identifiable Information (PII). The findings suggest that Decision Tree, Gradient Boosting, and Logistic Regression show higher predictive power even with limited non-PII-releasing sensor data. Certain combinations of non-PII releasing sensors can help achieve almost perfect performance for ML classifiers.
2023 IEEE INTERNATIONAL CONFERENCE ON DIGITAL HEALTH, ICDH
(2023)
Proceedings Paper
Computer Science, Theory & Methods
Dalyapraz Manatova, Inna Kouper, Sagar Samtani
Summary: This paper presents the development of a vulnerability management dashboard that assists security analysts in addressing challenges posed by network threats. Through interviews and real-world data, the typical workflow of security analysts is incorporated into the design, with vulnerability prioritization based on age, persistence, and impact. Future work includes user studies to evaluate the functionality and utility of the dashboard.
PRACTICE AND EXPERIENCE IN ADVANCED RESEARCH COMPUTING 2022
(2022)
Proceedings Paper
Computer Science, Interdisciplinary Applications
Dalyapraz Manatova, Dewesha Sharma, Sagar Samtani, L. Jean Camp
Summary: Underground markets provide a platform for merchants and buyers to trade assets using various digital currencies, payments providers, and wallets, which supports e-crime. Underground forums serve as critical intermediaries for marketplaces, enabling new entrants to establish trust and facilitating interactions between different markets. An empirical analysis of an underground forum helps understand online crime and reveals the network structure and interactions within different types of crimes.
2022 APWG SYMPOSIUM ON ELECTRONIC CRIME RESEARCH, ECRIME
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Agrim Sachdeva, Ben Lazarine, Ruchik Dama, Sagar Samtani, Hongyi Zhu
Summary: This research utilizes source code from GitHub to identify vulnerabilities in foundational repositories commonly used in modern AI development and AI repositories that depend on these foundational repositories. By using unsupervised graph embedding techniques, the study generates embeddings that capture vulnerability information and relationships between repositories, allowing for clustering of similarly vulnerable repositories. This research identifies patterns and similarities between repositories, which can aid in the development of effective vulnerability mitigation strategies for foundational AI repositories.
2022 IEEE INTERNATIONAL CONFERENCE ON DATA MINING WORKSHOPS, ICDMW
(2022)
Proceedings Paper
Computer Science, Information Systems
Murat Kantarcioglu, Barbara Carminati, Sagar Samtani, Sudip Mittal, Maanak Gupta
Summary: Modern civilization relies heavily on computing systems, which affect various aspects of business, government, and individual life. With the increasing emphasis on privacy and regulations, computing systems are expected to comply with these requirements, incorporating privacy and legal compliance into their design and being adaptable to changes.
CODASPY'22: PROCEEDINGS OF THE TWELVETH ACM CONFERENCE ON DATA AND APPLICATION SECURITY AND PRIVACY
(2022)
Article
Computer Science, Information Systems
Sagar Samtani, Yidong Chai, Hsinchun Chen
Summary: Black hat hackers cause significant economic losses worldwide, and cyber threat intelligence (CTI) helps organizations prioritize vulnerabilities. This study proposes a novel deep learning-based model (EVA-DSSM) and a device vulnerability severity metric (DVSM) to enhance CTI by analyzing the Dark Web. Evaluation results show that EVA-DSSM outperforms other methods in short text matching, and the CTI case studies demonstrate the utility of EVA-DSSM and DVSM for cybersecurity professionals.
Article
Computer Science, Information Systems
Mohammadreza Ebrahimi, Yidong Chai, Sagar Samtani, Hsinchun Chen
Summary: International dark web platforms contain a large amount of hacker assets, but analyzing them becomes challenging due to the lack of foreign-language training data. To address this issue, researchers have developed a novel technique for cross-lingual hacker asset detection, which leverages knowledge learned from English content to detect hacker assets in non-English dark web platforms. The study suggests that cybersecurity managers can benefit from focusing on Russian to identify sophisticated hacking assets, while financial hacker assets are scattered among several dominant dark web languages.
Article
Computer Science, Hardware & Architecture
Yunji Liang, Qiushi Wang, Kang Xiong, Xiaolong Zheng, Zhiwen Yu, Daniel Zeng
Summary: In this paper, a deep learning-based system called CyberLen is proposed for detecting malicious URLs effectively. The system utilizes factorization machine (FM) and position embedding to enhance the learning of interaction among URL features and reduce the ambiguity of URL tokens. Experimental results demonstrate the superior performance of the proposed method in terms of F1 score and convergence speed.
IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING
(2022)
