Article
Construction & Building Technology
Ruikai He, Tong Xiao, Shunian Qiu, Jiefan Gu, Minchen Wei, Peng Xu
Summary: This paper thoroughly analyzes the data quality issue of engineering big data from non-demonstration complexes in China and finds that the hourly power data of equipment groups are stable, the quality of pipe data is acceptable, the number of data types and the quality of cooling tower data are poor. The quality of other data is unstable. A rule-based data preprocessing framework is proposed, which utilizes the law of physics to ensure the strong coupling of multi-variants.
ENERGY AND BUILDINGS
(2022)
Article
Engineering, Civil
Hao-Han Xiao, Wen-Kun Yang, Jing Hu, Yun-Pei Zhang, Liu-Jie Jing, Zu-Yu Chen
Summary: This paper discusses the importance of preprocessing large amounts of data collected from tunnel boring machine excavations before using it for machine learning on TBM performance predictions. The research work is based on two water diversion tunneling projects and suggests using moving average methods and noise reduction filters to process the data. A drilling efficiency index is introduced to assess the relationships between mechanical parameters in a boring cycle. The paper also defines irrelevant data caused by human or mechanical errors and provides a program for recognizing and classifying these categories.
Review
Computer Science, Information Systems
Shaik Hasane Ahammad, Sandeep Dwarkanath, Rahul Joshi, B. T. P. Madhav, P. Poorna Priya, Osama S. Faragallah, Mahmoud M. A. Eid, Ahmed Nabih Zaki Rashed
Summary: In order to address the issues of cold-start and scalability in the collaborative filtering-based hotel recommendation system under a ranking list, this study proposed an extensive data analysis. Additionally, the user-friendly nature of the application was highlighted. This paper presents an innovative approach using Capsule Network (CapsNet) to recommend the most suitable hotel for the users. Overall, the importance of this research is rated at 7 out of 10.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Engineering, Multidisciplinary
Jing Zhang, Hongbo Liu, Xiaojun Sun, Shangyi Liu
Summary: This study proposed a tracking fusion Kalman filtering algorithm for processing building subsidence monitoring data, aiming to improve the accuracy of subsidence interpretation and prediction, providing a reference for building safety protection.
ALEXANDRIA ENGINEERING JOURNAL
(2021)
Article
Multidisciplinary Sciences
Hongyan Ma
Summary: This study aims to evaluate the applicability of the Distributed Clustering Algorithm in big data processing in power systems. A two-layer DCA algorithm based on K-Means Clustering and Affinity Propagation is proposed, combined with the Incentive Demand Response and a multi-period information economic dispatch model. Results show that the model can effectively consume new energy and meet the demand of the user side.
Article
Construction & Building Technology
Kris McGlinn, Rob Brennan, Christophe Debruyne, Alan Meehan, Lorraine McNerney, Eamonn Clinton, Philip Kelly, Declan O'Sullivan
Summary: BIM is crucial for integrating building data within the building life cycle, supporting various use cases related to automation, energy efficiency, and sustainability. Open building data faces challenges in standardization, data interdependency, access, security, and intellectual property protection. In Ireland, OSi maintains a large dataset called Prime2, including building GIS data and building specific data, supporting the development of a national geospatial identifier infrastructure based on an OSi building ontology.
AUTOMATION IN CONSTRUCTION
(2021)
Article
Multidisciplinary Sciences
Joshua T. Vogelstein, Eric W. Bridgeford, Minh Tang, Da Zheng, Christopher Douville, Randal Burns, Mauro Maggioni
Summary: Researchers have introduced a method that combines class-conditional moment estimates into low-dimensional projection, aiming to achieve more accurate dimensionality reduction for high-dimensional biomedical data for subsequent classification. The method has been validated on datasets with varying numbers of features, demonstrating improved accuracy and computational efficiency.
NATURE COMMUNICATIONS
(2021)
Article
Multidisciplinary Sciences
Halima Alachram, Hryhorii Chereda, Tim Beissbarth, Edgar Wingender, Philip Stegmaier
Summary: Biomedical and life science literature plays an important role in publishing experimental results, and the rapid growth of new publications has led to an increase in scientific knowledge represented in free text. Developing techniques to extract this knowledge using word2vec approach has shown to be effective in aiding scientists in discovering new relationships between biological entities. The study generated word vector representations based on a large corpus of PubMed abstracts, and demonstrated the utility of word2vec embeddings in biomedical analysis through validation experiments. By creating gene-gene networks and using them to train Graph-Convolutional Neural Networks, the study showed that word2vec-embedding-derived networks performed well in tasks such as predicting metastatic events in breast cancer, validating the usefulness of the generated word embeddings in constructing biological networks.
Review
Computer Science, Interdisciplinary Applications
Fangyu Li, Yuanjun Laili, Xuqiang Chen, Yihuai Lou, Chen Wang, Hongyan Yang, Xuejin Gao, Honggui Han
Summary: The construction industry is undergoing an intelligent revolution enabled by technologies like IoT, cloud computing, and robotics. Utilizing diverse big data from multiple sources can enhance efficiency, reduce waste and expenses, improve planning and decision-making processes, lower errors, and enhance safety at construction sites. This article provides a comprehensive review of the advantages and current state of big data in the construction industry, addressing unresolved difficulties and offering thoughts on its potential future.
JOURNAL OF INDUSTRIAL INFORMATION INTEGRATION
(2023)
Article
Computer Science, Information Systems
Changbo Ke, Fu Xiao, Zhiqiu Huang, Yunfei Meng, Yan Cao
Summary: In this paper, a method for private data chain disclosure discovery is proposed to prevent the illegal disclosure of a user's sensitive privacy information.
IEEE TRANSACTIONS ON SERVICES COMPUTING
(2022)
Article
Automation & Control Systems
Lan Yang, Kathryn Cormican, Ming Yu
Summary: Existing systems engineering standards are fragmented and communication between stakeholders is hindered due to discrepancies and misunderstandings. Transitioning from document-centric to model-based systems engineering requires advanced information exchange schema. An ontology learning methodology can automate the extraction of a sophisticated system engineering ontology to harmonize existing standards and improve interoperability.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Computer Science, Information Systems
Qi Li, Mingyu Cheng, Junfeng Wang, Bowen Sun
Summary: Phishing emails are becoming more complex, making existing detection methods inadequate. This article introduces an LSTM-based phishing detection method that achieves 95% accuracy through sample expansion and testing stages.
IEEE TRANSACTIONS ON BIG DATA
(2022)
Article
Computer Science, Interdisciplinary Applications
Simon Foll, Martin Maritsch, Federica Spinola, Varun Mishra, Filipe Barata, Tobias Kowatsch, Elgar Fleisch, Felix Wortmann
Summary: Researchers use wearable sensing data and machine learning models to predict health and behavioral outcomes, but data from commercial wearables often contain noise and artifacts. FLIRT is an open-source Python package that focuses on processing physiological data from commercial wearables, utilizing state-of-the-art algorithms for robust preprocessing and generating standardized feature vectors to improve reproducibility and performance in classification tasks.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2021)
Article
Computer Science, Information Systems
Da Zhang, Mansur R. Kabuka
Summary: The paper proposes a system infrastructure to construct a large knowledge graph of big scholar data, discover meta paths between entities, and calculate relevancy within the graph. This infrastructure utilizes Amazon EC2 for secured and private computing, processes data in parallel using Spark, and discovers relationships between entities distributedly. Additionally, four relationship discovery tasks are incorporated on top of this infrastructure using a mixed and weighted meta path (MWMP) method to explore potential relationships among different types of entities.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTING
(2021)
Article
Computer Science, Information Systems
Amrit Pal, Abishi Chowdhury, Satakshi, Husnu S. Narman, Arkabandhu Chowdhury, Manish Kumar
Summary: In this paper, a distributed classification technique for big data is presented, which efficiently utilizes distributed storage architecture and data processing units of a cluster. The proposed method does not require pre-structured data partitioning technique and is adaptive to big data mining tools. Extensive empirical analysis shows the effectiveness of the classifiers on benchmark datasets compared to other existing approaches.
Article
Computer Science, Artificial Intelligence
Sarah Shafqat, Maryyam Fayyaz, Hasan Ali Khattak, Muhammad Bilal, Shahid Khan, Osama Ishtiaq, Almas Abbasi, Farzana Shafqat, Waleed S. Alnumay, Pushpita Chatterjee
Summary: Healthcare Informatics is a phenomenon that has been discussed since the early 21st century. With the development of new computing technologies, a large amount of healthcare data is being produced, necessitating the management and extraction of knowledge for decision making. Researchers are exploring big data analytics, deep learning, predictive analytics, and other algorithms to bring innovation to healthcare. This research proposes a hybrid deep learning technique for medical diagnostics and tests and validates it using real-time datasets.
NEURAL PROCESSING LETTERS
(2023)
Article
Chemistry, Multidisciplinary
Sheeba Razzaq, Amil Roohani Dar, Munam Ali Shah, Hasan Ali Khattak, Ejaz Ahmed, Ahmed M. El-Sherbeeny, Seongkwan Mark Lee, Khaled Alkhaledi, Hafiz Tayyab Rauf
Summary: According to the World Health Organization, rear-ending collision is the leading cause of fatalities and injuries. This paper proposes a driver assistance system that analyzes accident contributing factors to improve vehicles' ability to avoid collisions.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Information Systems
Kamran Sattar Awaisi, Assad Abbas, Hasan Ali Khattak, Arsalan Ahmad, Mazhar Ali, Abbas Khalid
Summary: This paper proposes a Deep Reinforcement Learning-based framework for an IIoT-enabled smart parking system. By using smart cameras, fog nodes, and a cloud server, the system intelligently classifies vehicles and allocates parking slots. Experimental results show that the proposed method outperforms other techniques in terms of accuracy and processing time.
CLUSTER COMPUTING-THE JOURNAL OF NETWORKS SOFTWARE TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Kamran Ahmad Awan, Ikram Ud Din, Ahmad Almogren, Hasan Ali Khattak, Joel J. P. C. Rodrigues
Summary: The Internet of Things (IoT) is revolutionizing the world by making surrounding devices smart and capable of performing daily-life activities with precision. IoT and healthcare collaborate to provide notable facilities in patient monitoring. However, the identification of malicious and compromised nodes remains a critical challenge. This article proposes a machine learning-based trust management approach for edge nodes to detect nodes with malicious behavior. The approach utilizes knowledge and experience components of trust and collects recommendations from edge clouds to evaluate indirect and aggregated trust.
Article
Computer Science, Information Systems
Zainab Iftikhar, Adeel Anjum, Abid Khan, Munam Ali Shah, Gwanggil Joen
Summary: With the growth of VANET technology, data generated by communication among vehicular devices and edge nodes becomes massive. Privacy preservation is a major challenge due to the personal and sensitive information contained in the data. Most existing distributed privacy preserving solutions rely on third-party anonymization, but Local Differential Privacy (LDP) allows for local and individual data anonymization without a third party. In this work, a privacy preservation solution using LDP is proposed to address security and privacy threats. Additionally, a model incorporating IOTA ledger further enhances privacy and security in a complex and distributed network of vehicles.
CLUSTER COMPUTING-THE JOURNAL OF NETWORKS SOFTWARE TOOLS AND APPLICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Basker Palaniswamy, Keyvan Ansari, Alavalapati Goutham Reddy, Ashok Kumar Das, Sachin Shetty
Summary: This article discusses the challenges of authentication in controller area network (CAN) buses within an intra-vehicular network involving electronic control units (ECUs). The existing comprehensive protocol suite is formally analyzed, and two new authentication protocols are proposed to mitigate common attacks.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2023)
Article
Engineering, Electrical & Electronic
Maria Naz, Munam Ali Shah, Hasan Ali Khattak, Abdul Wahid, Muhammad Nabeel Asghar, Hafiz Tayyab Rauf, Muhammad Attique Khan, Zoobia Ameer
Summary: Pandemics and natural disasters are increasing, causing more pressure on life care services and users. There is a lack of knowledge on how to prevent these disasters and pandemics. In this study, a model with 12 branches of CNN was proposed to detect different diseases and their subtypes using CT scan images, achieving accurate classification.
INTERNATIONAL JOURNAL OF IMAGING SYSTEMS AND TECHNOLOGY
(2023)
Article
Chemistry, Multidisciplinary
Bello Musa Yakubu, Majid Iqbal Khan, Abid Khan, Adeel Anjum, Madiha Haider Syed, Semeen Rehman
Summary: Advancements in sensor-enabled devices led to the emergence of resource trading models for smart communities, such as the smart marketplace (SMP). Most of the proposed SMP architectures are based on blockchain technology, which has a public ledger to achieve transparency. However, safeguarding the participant's anonymity, untraceability, and transactional data privacy during trading becomes a challenging task.
APPLIED SCIENCES-BASEL
(2023)
Article
Medicine, General & Internal
Muhammad Asif, Munam Ali Shah, Hasan Ali Khattak, Shafaq Mussadiq, Ejaz Ahmed, Emad Abouel Nasr, Hafiz Tayyab Rauf
Summary: Intracranial hemorrhage (ICH) requires immediate action from radiologists as it can cause death or disability. Existing artificial intelligence methods for ICH detection and subtype classification lack accuracy. In this paper, a new methodology called ResNet101-V2, Inception-V4, and LGBM (Res-Inc-LGBM) is proposed, which achieves high accuracy, sensitivity, and F1 score for ICH detection and subtype classification using brain CT scans. The proposed solution outperforms standard benchmarks and shows the significance of its real-time application.
Article
Computer Science, Hardware & Architecture
Waseeq Ul Islam Zafar, Muhammad Atif Ur Rehman, Farhana Jabeen, Rehmat Ullah, Ghulam Abbas, Abid Khan
Summary: This paper discusses the importance of inter-vehicle communication in VANETs and the use of NDN as the underlying protocol. It proposes a decentralized receiver-based link stability-aware forwarding (DRLSF) protocol to address the challenges faced. The DRLSF protocol is suitable for pull-based applications.
Review
Chemistry, Analytical
Mansoor Ahmed, Amil Rohani Dar, Markus Helfert, Abid Khan, Jungsuk Kim
Summary: Data provenance is a method of recording data origins and the history of data generation and processing. In healthcare, it is important to implement data provenance to track the sources and reasons behind any issues with user data. This systematic review explores the impacts of data provenance in healthcare and GDPR-compliance-based data provenance, discussing the technologies and methodologies used to achieve it. The study identifies research gaps and suggests future directions.
Article
Chemistry, Analytical
Muhammad Irfan Khalid, Mansoor Ahmed, Jungsuk Kim
Summary: Dynamic consent management systems allow individuals to dynamically control access to their data. Security and privacy guarantees are crucial for the adoption of such systems, with specific data protection requirements needed for compliance with regulations like the GDPR. This paper explores data protection issues in dynamic consent management systems, identifying key security and privacy properties and discussing limitations in existing systems. It proposes using tools and technologies like differential privacy, blockchain, zero-knowledge proofs, and cryptogrpahic procedures to build secure and private dynamic consent management systems.
Article
Computer Science, Theory & Methods
Hira S. Sikandar, Saif ur Rehman Malik, Adeel Anjum, Abid Khan, Gwanggil Jeon
Summary: Federated Learning (FL) is a decentralized machine learning strategy where clients locally train on a shared global model provided by the server. However, the widespread adoption of FL in distributed settings has led to security attacks, including label-flipping attacks. This research proposes a defense mechanism based on Type-based Cohorts (TC) with Kernel Principal Component Analysis (KPCA) to detect and defend against such attacks. Additionally, Multi-path Service Routing (MSR) is deployed to improve network performance.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2023)
Article
Green & Sustainable Science & Technology
Azhar Mahmood, Abid Khan, Adeel Anjum, Carsten Maple, Gwanggil Jeon
Summary: Smart Grids have several advantages over traditional grids, but they also introduce security and privacy issues. Data aggregation plays a crucial role in protecting user consumption data, but existing schemes have limitations. This paper proposes a decentralized secure data aggregation scheme using blockchain to preserve the privacy, integrity, and authentication of individual consumption data.
SUSTAINABLE ENERGY TECHNOLOGIES AND ASSESSMENTS
(2023)