Prof. Hong Lin, University of Houston-Downtown, USA
Speech Title: Sleep stage analysis by Bayesian Network
Abstract: Our research team built a self-learning Bayesian network adept at elucidating intricate relationship mappings within raw single-channel electroencephalography (EEG) data. The algorithm employed probabilistic graphical models to decode both the dynamic and static interdependencies inherent in EEG signals, providing an unprecedented framework for an in-depth understanding of sleep dynamics.
By proper signal discretization and construction of a Bayesian network, the system can autonomously identify and intricately model both time-invariant features and their dynamic relationships across various sleep stages. The innovative methodology we employed focused upon uncovering relationships present in the data by pivoting around the construction of a Bayesian network which autonomously learns probabilistic relationships from EEG data. Our approach diverges radically from traditional methods by seamlessly integrating automated feature extraction and intricate temporal sequence analysis within a single, unified modeling framework.
Biography: Hong Lin received his PhD in Computer Science from the University of Science and Technology of China. Before he joined the University of Houston-Downtown (UHD), he was a postdoctoral research associate at Purdue University, and an assistant research officer at the National Research Council, Canada. Dr. Lin is currently a Professor in Computer Science with UHD. His research interests include cognitive intelligence, human-centered computing, parallel/distributed computing, and big data analytics. He is the supervisor of the Grid Computing Lab and a co-founder of the Data Center at UHD. Dr. Lin currently serves as the program director for the Master of Science in Artificial Intelligence program at UHD. Dr. Lin is a senior member of the Association for Computing Machinery (ACM).
Prof. Koen Smi, HU University of Applied Sciences Utrecht, Netherlands
Speech Title: Applied Research into Digital Twinning to Support the Policy Lifecycle and Spatial Planning by Regional and Local Governments in the Netherlands
Abstract: An increasing number of Dutch provinces and municipalities are engaged in the experimentation of Digital Twin technology, which involves the creation of digital replicas of regions, areas, cities, and neighborhoods. The ultimate objective is to support policy development and spatial planning practices. Digital Twins hold significant promise due to their capacity to incorporate and combine various areas of interest, such as biodiversity, mobility, heat stress, flood risk, sound levels, and more, within a dynamic digital (3D) environment. Digital Twins also enable real-time manipulation of variables pertaining to these areas of interest.
At present, numerous independent initiatives involving Digital Twin technology are being pursued by Dutch provinces and municipalities; however, a lack of adequate collaboration exists among them. The Digital Twins lab at HU University of Applied Sciences Utrecht aims to build and facilitate extensive cooperation among these Dutch provinces and municipalities. To achieve this, the lab conducts research on various aspects, including legal, technical, governance, and ethical considerations.
This presentation will primarily focus on the structure and functioning of the Digital Twins lab, the outcomes attained through our studies, and the valuable insights gained from multiple (applied) research studies.
Biography: Koen Smit is a professor focusing on Digital Ethics at the HU University of Applied Sciences Utrecht, in the Netherlands. He obtained his PhD in Computer Science in 2018 at the Open Universiteit. His research primarily focuses on the combination of Business Process Management, Business Rules Management, Decision Management, Decision Mining, Digital Twin technology, and Social Robotics and Value Sensitive Design. His interest also leans towards how said technological innovations can be designed and implemented in such a way that human and public values are explicitly and adequately considered. He regularly reviews and/or publishes and presents his research contributions at conferences and journals (e.g., HICSS, ICIS, PACIS, AMCIS, PJAIS, JITTA, and BPM). Furhtermore, he is part of the management team of the Institute for ICT of the same university. He supervises several PhD and Professional Doctorate students on his focus areas.
Prof. Ghulam Abbas, GIK Institute of Engineering Sciences and Technology, Pakistan
Speech Title: Instruction-Level Parallelism in High-Performance and Networked Systems
Abstract: As computational demands grow, optimizing processor efficiency is essential for high-performance and networked computing environments. Instruction-Level Parallelism (ILP) plays a key role in improving execution speed by enabling multiple instructions to be processed concurrently. This talk explores ILP fundamentals, covering pipelining, multiple issue, and out-of-order execution techniques to enhance processor throughput.
We will discuss the impact of ILP on modern computing, particularly in data centres, HPC clusters, and AI-driven systems, where efficient instruction processing is critical. Additionally, we will examine pipeline hazards and strategies to mitigate them. The session will also highlight advanced ILP techniques and recent trends, which power modern high-speed processors.
By the end of this talk, attendees will gain a deeper understanding of how ILP optimizations contribute to computational efficiency in networked and high-performance systems, making it a crucial concept for both academia and industry.
Biography: GHULAM ABBAS received the B.S. degree in computer science from University of Peshawar,Pakistan, in 2003, and the M.S. degree in distributed systems and the Ph.D. degree in computer networksfrom University of Liverpool, U.K., in 2005 and 2010, respectively. From 2006 to 2010, he was ResearchAssociate with Liverpool Hope University, U.K., where he was associated with the Intelligent & DistributedSystems Laboratory. Since 2011, he has been with the Faculty of Computer Science & Engineering, GIKInstitute of Engineering Sciences and Technology, Pakistan. He is currently working as a full Professor,Head of Cybersecurity and Software Engineering Departments, and Director ICT Academy. Dr. Abbas is aco-founding member of the Telecommunications and Networking (TeleCoN) Research Center at GIKInstitute. He is a Fellow of the Institute of Science & Technology, U.K., a Fellow of the British ComputerSociety, and a Senior Member of the IEEE. His research interests include computer networks and wirelessand mobile communications.
Prof. Paulo Batista, University of Évora, Portugal
Speech Title: Theoretical Information Science
Abstract: Following the Second World War an explosion in the quantity of documentation led to a dramatic change in Archiving, or the profession referred to as records managers/records management and archivists/archives. Starting in the 1980s, however, archivists in Quebec began to make great progress by changing their approach and looking at the entire documentary cycle from current to definitive information. Carol Couture and Jean- Yves Rousseau made a crucial contribution towards the understanding of the Three Age Theory that viewed Archiving as an integrated discipline centered on a structural understanding of archives. In 1994, their work Les Fondements de la Discipline Archivistique, presented a new interpretation of Theodore Schellenberg's Three Age Theory. They called attention to the fact that the three phases of archival documents are not separate but, on the contrary, integrated. They argued that these three stages can even be looked at in a segmented way, provided the union between them is ensured. Their great innovation relative to Schellenberg's work lay, precisely, in critiquing the division and separation between the three ages of archival documents. Couture and Rousseau thereby brought together all the phases of the lifecycle of records, from production to dissemination, in opposition to the sterile distinction advocated by traditional archivists and document managers. In my opinion, however, the best approach to integrating information management is known as records continuum, which places archives in a post-custodial, informational, and scientific paradigm. This Australian concept arose in the 1990s amid the huge explosion of information, communication technologies and new media. This context forced Information Science to redefine its object of study. Records continuum is closely related to the integrated management model of Couture and Rousseau, while it carries their innovation further, perfecting it and replacing it with systemic dynamics and providing continuity between archives. In fact, records continuum means, literally, continuous management. It looks at the whole process from the production of records to their final archiving. Otherwise, we cannot speak of continuous management. That is why, when we speak of rigid archives – current, intermediate, and definitive, this approach is more theoretical than practical. There is, in fact, no separation between these phases, even less so from the point of view of the value of documents. The traditional distinction between information with probative and historical value ceases to exist. The information is simultaneous and is, in fact, the same.
Biography: Paulo Batista is PhD Researcher at CIDEHUS.UÉ-Interdisciplinary Center for History, Culturesand Societies of the University of Évora, Portugal, where is the coordinator of the research group2:Heritage and Literacies. Currently works as professor at the Is cte-IUL, in the Master in Architecture andVisual Culture in Lisbon, and at the Autonomous University of Lisbon, where is coordinator and professorof the Postgraduate in Promotion and Cultural and Educational Dynamization of Archives and Libraries,and the Postgraduate in Architectural Archives.
Prof. Jyotsna Kumar Mandal, University of Kalyani, India
Speech Title: Automated Confidence Based Learning and Assessment
Abstract: In this lecture the requirements for development of a system that will promote CBL with an architecture of the same. Existing methods, techniques, and models available in the field of e-learning or TEL are studied and presented and the research gap has been formulated. Moreover. The talk stressed upon the requirement of an enhanced learning content development framework and augmented assessment methodology. In proceeding through the lecture it became evident that as the CBL is concerned with the world of work, input from the jobsite is of immense importance in establishing the performances and adequacies of the learners. Learning Record Store (LRS) are used to identify the adequacies and in promoting the customized content development. The customization of the content and prescribing it to the learner is also important and a method to be discussed in this area. There is a requirement of monitoring the performance of learning method will be discussed.
Biography: Dr. Jyotsna Kumar Mandal, M. Tech. in Computer Science from University of Calcutta in 1987, awarded Ph. D. (Engineering) in Computer Science and Engineering by Jadavpur University in 2000. Working as Professor of Computer Science & Engineering, University of Kalyani. Former Vice Chancellor, Raiganj University, West Bengal, Former Dean, Faculty of Engineering, Technology & Management, KU for two consecutive terms during 2008-2012. Former Director, IQAC and Chairman CIRM Kalyani University. Served as Professor Computer Applications, Kalyani Government Engineering College for two years. He was Associate Professor Computer Science for eight years at North Bengal University and Assistant Professor Computer Science North Bengal University for seven years. He also served as lecturer at NERIST, Itanagar for one year. 36 years of teaching and research experience in Coding Theory, Data and Network Security and authentication; Remote Sensing & GIS based Applications, Data Compression, Error Correction, Visual Cryptography and Steganography. Awarded 30 Ph. D. Degrees and 8 are pursuing. Supervised 03 M. Phil, more than 80 M. Tech and more than 130 M.C.A. Dissertations. Published more than 450 research articles. Recently he has published a text book on Reversible Steganography and Authentication via Transform Encoding from Springer (https://link.springer.com/book/10.1007/978-981-15-4397-5). This book has been translated into Chinese and republished from China by Springer. Organized more than 60 International Conferences and Corresponding Editors of edited volumes and conference publications of Springer, IEEE and Elsevier etc. and edited 60 volumes as volume editor. Received "Shiksha Ratna" Award from Government of West Bengal, India for outstanding teaching and research work in 2018. ISO world Convenor of ISO/IEC JTC 1/SC36/WG7. Governing Council(GC) Member of IETE, India.
Prof. Chuan-Ming Liu, National Taipei University of Technology, Taiwan
Speech Title: Learned Indices for Spatial Data
Abstract: An index is a structure or organization on data for effectively managing data item in terms of time and space, such as hash tables, binary search trees, and B-trees. As the properties and types of data change over time, new appropriate indices for efficient management on data become more and more important and necessary. On the other hand, as the techniques of machine learning or deep learning advance, many applications using machine learning for a better performance have been explored. Recall the idea and objective of an index. The index now can be seen as a model in machine learning, which can locate the data item effectively by prediction. With this observation, a learned index, a model that considers the patterns and distributions of data, has been proposed to facilitate search processing. Some learned indices have been provided for one-dimensional data, including Range Index and Recursive Model Index (RMI). For multi-dimensional (or spatial) data, it is always a challenging work to have effective index structures. Some well-known spatial indices, like kd-trees, quad-trees, and R-trees, with their variants for improvement on the efficiency have been studied till now. It thus is interesting and worthy to study the learned indices on spatial (multi-dimensional) data for a better performance. In this talk, the learned indices will be introduced starting with the ones for one-dimensional data. We then focus on the learned indices for spatial data and present our learned indices based on index tree structures. With the learned indices as models, evaluation on preprocessing, training, prediction, error, as well as query processing for point, range and kNN queries will be addressed as well.
Biography: Dr. Chuan-Ming Liu is a professor in the Department of Computer Science and Information Engineering (CSIE), National Taipei University of Technology (Taipei Tech), TAIWAN, where he was the Department Chair from 2013-2017. He received his Ph.D. in Computer Science from Purdue University in 2002 and joined the CSIE Department in Taipei Tech in the spring of 2003. In 2010 and 2011, he has held visiting appointments with Auburn University, Auburn, AL, USA, and the Beijing Institute of Technology, Beijing, China. He has services in many journals, conferences and societies as well as published more than 120 papers in many prestigious journals and international conferences. Dr. Liu was also the co-recipients of the best paper awards in many conferences, including ICUFN 2015, ICS 2016, MC 2017, WOCC 2018, MC 2019, WOCC 2021, TCSE 2022, and TANET 2023. His current research interests include data science, big data management, uncertain data management, spatial data processing, data streams, ad-hoc and sensor networks, location-based services.
Prof. Yingwah Teh, University of Malaya, Malaysia
Speech Title: Modern Data Mining in Information Systems: A 2025 Perspective
Abstract: This talk examines the role of modern data mining techniques in advancing information systems, with a focus on enhancing property developer experiences. As data generation grows exponentially, data mining has become critical for optimizing systems, improving security, and enabling intelligent, real-time decision-making. Key techniques such as machine learning, clustering, and anomaly detection are applied to improve resource allocation and system efficiency in distributed and cloud-based environments. We also share our experience in developing data-driven solutions for the property development sector, addressing challenges like data privacy and scalability while exploring trends such as AI-driven analytics and edge computing.
Biography: As a highly accomplished computer scientist and data mining expert with over 35 years of experience, he has demonstrated exceptional leadership, expertise, and vision in the field.
Over the course of my career, he has achieved numerous successes and made significant contributions to the industry. He began as an entry-level computer programmer in 1988 and advanced to become a Professor of Data Mining at the Faculty of Computer Science and Information Technology at the University of Malaya. He obtained my tertiary academic qualifications from Oklahoma City University and the University of Malaya, and he has published more than 90 academic papers in top-tier journals, including Information Fusion and the International Journal of Information Management.
He has a remarkable H-index and number of citations in Web of Science, Scopus, and Google Scholars databases, and he has supervised numerous students at all levels of study. His areas of research include data warehouse, data mining, deep learning, IoT, activity recognition, wearable sensors, accelerometers, heart arrhythmia, electrocardiograph, supraventricular premature beat, multivariate time series, edge computing, task scheduling, data streams, mobile computing, speaker verification, language recognition, clustering algorithms, MapReduce, stock market, and sentiment analysis.
He has received several grants of more than RM one million, including public, international, and private grants, and he has completed two commercial data science projects for Petronas GTD and Air Liquide. I serve as an Associate Editor for Human-Media Interaction – Frontiers in Psychology and a reviewer for several high-quality journals. I am also an Expert Advisory Panel for Master of Science (Data Science) Degree Program at UTP, Programme Advisory Panel for Bachelor of Business (Honors) in Business Analytics at TARUC, and an external
assessor of Swinburne University of Technology (BS of Computer Science program). Additionally, he is an External Assessor for Programme Master Sciences (Computer and Information Engineering), IIUM, and a technical assessor of Swiss National Science Foundation.
He has been teaching data mining since 2002, and has produced many highly qualified data scientists and Ph.D. graduates who have gone on to work for top companies like IBM, Amazon Web Services, and Google.
As a highly respected and accomplished computer scientist, he has demonstrated exceptional dedication to the industry, and I am confident that my expertise, leadership, and vision make me a highly qualified computer scientist.
Assoc. Prof. Rohaya Binti Latip, Universiti Putra Malaysia, Malaysia
Speech Title: Task Offloading Optimization Algorithm for Time-Energy Minimization in Mobile Edge Computing
Abstract: Mobile edge computing (MEC) is a promising technology that enhances computational capacity and reduce the latency at the edge of mobile networks, yet task offloading in edge computing remains challenging due to its dynamic configurations and the significant resource requirements involved. Existing literature has made strides in addressing these issues; however, many solutions fail to adequately consider the optimal selection of tasks and their inherent dependencies, which are crucial for efficient offloading. Therefore, this paper proposed a weighted quantum particle swarm and dispatched task offloading problem in mobile edge computing as a multi-objective function aimed at minimizing both energy consumption and task completion time. We introduce a novel weighted quantum particle swarm optimization with Dispatched (WQPSOD) task offloading optimization algorithm designed to enhance performance in mobile edge computing environments. The WQPSOD algorithm is implemented in Python and evaluated in a multi-user, multi-server context. Our experimental results demonstrate that WQPSOD significantly outperforms benchmark algorithms, by achieving 60.88% and 51.10% improvement in task completion time and system energy consumption respectively.
Biography: She is currently an Associate Professor at the Department of Communication Technology and Network, Faculty of Computer Science and Information Technology, Universiti Putra Malaysia. She holds a Ph. D in Distributed Database and Msc. in Distributed System from Universiti Putra Malaysia. She graduated her Bachelor of Computer Science from University Technology Malaysia, Malaysia. She was the head of Department of Communication Technology and Network from 2017 until 2022. She served as an Associate Professor at Najran university, Kingdom of Arab Saudi from 2012 until 2013. She was the Head of HPC section in Universiti Putra Malaysia (2011-2012) and consulted the Campus Grid project and the Wireless for hostel in Campus UPM project. Her research interests include Big Data, Cloud, Fog and Edge Computing, Network management, and Distributed database. She has published more than 140 papers in international and national journals, proceedings and posters.
Assoc. Prof. Teoh Ai Ping, Universiti Sains Malaysia, Malaysia
Speech Title: Predicting Cybersecurity Behaviour among Professionals in Malaysia
Abstract: This paper explores the factors influencing cybersecurity behavior among accounting professionals in Malaysia by integrating Protection Motivation Theory (PMT) and Technology Threat Avoidance Theory (TTAT). Using a survey-based approach, data were gathered from 285 professionals through an online questionnaire, with 260 valid responses analyzed using SmartPLS. The findings reveal that perceived severity and rewards significantly impact perceived threat, whereas perceived vulnerability does not show a direct effect. Protection motivation is directly influenced by perceived threat and self-efficacy, while response cost does not have a significant impact. Moreover, protection motivation serves as a key mediator between perceived threat and self-efficacy in shaping cybersecurity behavior but does not mediate response cost. Cybersecurity awareness moderates the impact of response cost but not perceived threat or self-efficacy, whereas safeguard effectiveness negatively moderates the relationship between protection motivation and cybersecurity behavior. This study enriches the existing literature by examining the interplay between perceived threat, response cost, self-efficacy, protection motivation, and cybersecurity behavior, offering valuable insights for researchers, practitioners, and policymakers to enhance cybersecurity practices in professional settings.
Biography: Associate Professor Ts. Dr. Teoh Ai Ping currently serves as the Deputy Dean (Research, Innovation, Industry Community Engagement) at the Graduate School of Business, Universiti Sains Malaysia. She holds a Doctor of Business Administration, Master of Science (Information Technology) and Bachelor of Accountancy (Hons.). Ts. Dr. Teoh is a Professional Technologist (Cyber Security Technology) with the Malaysian Board of Technologists and a Certified Risk and Compliance Management Professional with the International Association of Risk and Compliance Management Professionals. She is also a member with the Malaysian Institute of Accountants, the Institute of Internal Audit, Malaysia and Association of Certified Fraud Examiners and member with Association of Computing Machinery and Institute of Electrical and Electronic Engineers. With a background in both accounting and information systems technology sector, she began her career as an account executive in the EMS industry and then transitioning to consulting in SAP R/3 Enterprise Resource Planning system specializing in the Financial and Controlling; as well as ABAP/4 programming and SAP Scripts. Ts. Dr. Teoh eventually embarked on a career in the academic sector as a Deputy Dean School of Business Administration in a private University before joining GSB USM. Ts. Dr. Teoh has served as external assessor and subject matter expert in several accounting and business programs and being external examiner for Doctoral theses evaluation. Ts Dr. Teoh has successfully completed several consultancy and corporate training projects to clients in the areas related to Business Information Systems in Malaysia and abroad. She has a widespread publication record, with articles featured in both local and international journals; and delivered speeches at various international conferences and industry events as Invited Speaker and Keynote Speaker. Her areas of interests include enterprise information systems, enterprise risk management systems, cyber security, and business sustainability.
Assoc. Prof. Tze Wei Liew, Multimedia University Malaysia
Speech Title: A Socio-Communicative Approach to Designing AI Systems
Abstract: As artificial intelligence (AI) systems become increasingly integrated into everyday interactions, their ability to engage users socially and communicate effectively is crucial. In this session, I will discuss the role of social cues in AI design, drawing from insights in human-agent interaction, communication theories, and the computers-are-social-actors paradigm. I propose a model that outlines how demographics, appearance, social prestige, specialization, communication style, and information quality (DASSCI model) shape users’ perceptions of AI competence and trustworthiness. By strategically embedding these socio-communicative elements, AI systems—whether embodied virtual agents, chatbots, or voice assistants—can enhance engagement, credibility, and user acceptance. This presentation will explore socio-communicative models for designing socially intelligent AI and its implications across education, healthcare, and e-commerce. The session aims to bridge theoretical frameworks with practical AI design strategies, offering insights for researchers and developers working on next-generation AI systems.
Biography: Tze Wei Liew is an Associate Professor of Information Science attached to the Centre for Interaction and Experience Design and the Faculty of Business at Multimedia University (MMU), Malaysia, specializing in teaching information systems and sciences to undergraduate and postgraduate students. His research interests and contributions in WoS and Scopus-indexed publications focus on human-media and human-agent interaction, with an emphasis on educational, instructional, media, and cyberpsychology. A member of the Association for Computing Machinery (ACM), he also serves on the editorial boards of journals, including Elsevier's Learning and Instruction and Wiley's Human Behavior and Emerging Technologies. He has actively collaborated on research presentations, lectures, and activities at international scholarly conferences and venues such as Australia, China, Cambodia, Denmark, India, Indonesia, Japan, Portugal, Singapore, Spain, Sweden, Taiwan, Thailand, and Vietnam, while serving as a Technical Program Committee (TPC) member and program chair for ACM and IEEE conferences in information sciences.
Dr. Chiagoziem Chima Ukwuoma, Chengdu University of Technology, China
Speech Title: Enhancing Solar GHI Forecasting with Dual-Input Features and Lightweight Transformer Models
Abstract: Accurate Global Horizontal Irradiance (GHI) forecasting is essential for efficient solar energy management. Traditional statistical models such as ARIMA and numerical weather prediction (NWP) models, though interpretable, struggle with capturing the nonlinear dependencies and rapid weather variations affecting solar radiation. Deep learning models, including Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks, have improved forecasting by leveraging spatial and temporal patterns. However, CNNs lack temporal awareness, while LSTMs suffer from high training costs and vanishing gradient issues. Transformer models have recently emerged as a powerful alternative, effectively capturing long-range dependencies through self-attention mechanisms. However, their high computational complexity limits their practicality in real-time applications. Another major challenge in GHI forecasting is data modality. Numerical weather data, such as temperature, humidity, and wind speed, provide useful meteorological insights but lack spatial granularity. Conversely, image-based models utilizing sky or satellite images offer spatial awareness but demand extensive storage and computational resources. Relying solely on either numerical or image-based inputs restricts forecasting accuracy and robustness.
We propose the design and application of dual-input lightweight transformer models that can integrate numerical weather data with sky image features to enhance forecasting accuracy while minimizing computational overhead. A CNN encoder extracts spatial features from sky images, while numerical weather features are processed through a dedicated transformer branch. The model employs a self-attention mechanism to effectively combine both data sources, capturing intricate spatial-temporal dependencies. Experimental results demonstrate that the proposed approach outperforms statistical models, standalone CNNs, and conventional transformer models in short- and medium-term GHI forecasting. The fusion of numerical and image-based data significantly improves predictive accuracy, while the lightweight transformer architecture ensures computational efficiency, making it viable for real-time deployment in solar energy management systems. This study underscores the importance of hybrid deep learning architectures in advancing solar forecasting, contributing to a more stable and efficient renewable energy grid.
Biography: Dr. Chima holds a Bachelor of Engineering in Mechanical Engineering (Automotive Technology) from the Federal University of Technology Owerri (FUTO), Nigeria, a Master of Science in Software Engineering, and a Doctor of Philosophy in Software Engineering, both from the University of Electronics Science and Technology of China (UESTC). He is currently a Senior Lecturer and Senior Researcher at OBU, Sino-British Collaborative Education, Chengdu University of Technology, China. Dr. Chima has published over 70 peer-reviewed papers in Top journals including Journal of Applied Energy, Renewable Energy, Biomedical Signal Processing and Control, International Journal of Hydrogen Energy, Computers in Biology and Medicine, Social Science and Medicine, Science of Total Environment, Biocybernetics and Biomedical Engineering, Clean Energy, Expert Systems and Application and lots more. He serves as an academic judge for the United States Academic Decathlon & Pentathlon (USAD & USAP) China and the National Economics Challenge (NEC) since 2019 and has been recognized with several prestigious awards, including the Centre for West African Studies of UESTC Doctoral Research Fund.
Jackey Cheung, The Chinese University of Hong Kong, Hong Kong, China
Speech Title: An adaptive auto-tuning Healthcare system development in Hyperledger Besu Blockchain
Abstract: Blockchain (BC) is widely regarded as one of the most groundbreaking technologies in this decade, distinguished by its key attributes of decentralization, security, and accessibility. In this presentation, we aim to share our insights and experiences in examining the performance characteristics of Blockchain applications within Healthcare IoT (IoHT). Our focus will be on critical metrics such as transaction throughput, latency, and resource utilization.
We will also discuss our approach to designing comparative experiments, considering parameters such as transaction send rate, block size, consensus mechanisms, and block time; and our investigation of the proof of authority consensus algorithms—namely QBFT, IBFT 2.0, and Clique, which was conducted using Hyperledger Caliper. We will analyze how these parameters influence the performance of a private Hyperledger Besu blockchain.
Building on our findings, we have delved into the development of our proposed Hyperledger Besu auto-tuning system, which employs a tunnel-limiter to guide the system toward optimal operational conditions. This leads us to our adaptive BC-parameter-tunable Decentralized framework for IoHT (ABCD-IoHT), designed to enhance the throughput performance of medical healthcare systems while ensuring robust security under varying medical load conditions.
Biography: Jackey Cheung is affiliated with the Department of Computer Science and Engineering. He has 15+ years teaching experience in CS/IT area in Universities; and he has taught many courses such as Blockchain, Artificial Intelligent, Cyber Security, Machine Learning, Computer Programming, Data Structures and Algorithms, Database Systems, Information Retrieval, Computer Graphics, Software Engineering, Software Management, Digital Literacy and Computational Thinking, etc. in University UG-Level and Master-Level.
His research interests: blockchain, artificial intelligent, cyber security, deep learning, virtual reality, computer vision, etc.
He likes sports (University-team-member), photography (advisor with photo-albums), traveling (exploring places like Iceland, Yellowknife, etc.), scuba-diving (advanced-diver), windsurfing, k-boating, archery, driving (in-training-pilot), reading (esp. history), HiFi-music, strategy-gaming, programming, and making new things [Welcome to join me if you have similar interest~]