Recent Submissions

  • Enhancing text comprehension via fusing pre-trained language model with knowledge graph

    Qian, Jing; Li, Gangmin; Atkinson, Katie; Yue, Yong; Xi'an Jiaotong-Liverpool University; University of Bedfordshire; University of Liverpool (Association for Computing Machinery, 2024-02-16)
    Pre-trained language models (PLMs) such as BERT and GPTs capture rich linguistic and syntactic knowledge from pre-training over large-scale text corpora, which can be further fine-tuned for specific downstream tasks. However, these models still have limitations as they rely on knowledge gained from plain text and ignore structured knowledge such as knowledge graphs (KGs). Recently, there has been a growing trend of explicitly integrating KGs into PLMs to improve their performance. For instance, K-BERT incorporates KG triples as domain-specific supplements into input sentences. Nevertheless, we have observed that such methods do not consider the semantic relevance between the introduced knowledge and the original input sentence, leading to the issue of knowledge impurities. To address this issue, we propose a semantic matching-based approach that enriches the input text with knowledge extracted from an external KG. The architecture of our model comprises three components: the knowledge retriever (KR), the knowledge injector (KI), and the knowledge aggregator (KA). The KR, built upon the sentence representation learning model (i.e. CoSENT), retrieves triples with high semantic relevance to the input sentence from an external KG to alleviate the issue of knowledge impurities. The KI then integrates the retrieved triples from the KR into the input text by converting the original sentence into a knowledge tree with multiple branches, the knowledge tree is transformed into an accessible sequence of text that can be fed into the KA. Finally, the KA takes the flattened knowledge tree and passes it through an embedding layer and a masked Transformer encoder. We conducted extensive evaluations on eight datasets covering five text comprehension tasks, and the experimental results demonstrate that our approach exhibits competitive advantages over popular knowledge-enhanced PLMs such as K-BERT and ERNIE.
  • Revolutionising financial portfolio management: the non-stationary transformer's fusion of macroeconomic indicators and sentiment analysis in a deep reinforcement learning framework

    Liu, Yuchen; Mikriukov, Daniil; Tjahyadi, Owen Christopher; Li, Gangmin; Payne, Terry R.; Yue, Yong; Siddique, Kamran; Man, Ka Lok; Xi’an Jiaotong-Liverpool University; University of Liverpool; et al. (MDPI, 2023-12-28)
    In the evolving landscape of portfolio management (PM), the fusion of advanced machine learning techniques with traditional financial methodologies has opened new avenues for innovation. Our study introduces a cutting-edge model combining deep reinforcement learning (DRL) with a non-stationary transformer architecture. This model is designed to decode complex patterns in financial time-series data, enhancing portfolio management strategies with deeper insights and robustness. It effectively tackles the challenges of data heterogeneity and market uncertainty, key obstacles in PM. Our approach integrates key macroeconomic indicators and targeted news sentiment analysis into its framework, capturing a comprehensive picture of market dynamics. This amalgamation of varied data types addresses the multifaceted nature of financial markets, enhancing the model’s ability to navigate the complexities of asset management. Rigorous testing demonstrates the model’s efficacy, highlighting the benefits of blending diverse data sources and sophisticated algorithmic approaches in mastering the nuances of PM.
  • A Big Data maturity model for Electronic Health Records in hospitals

    Daraghmeh, Rania; Brown, Raymond; University of Bedfordshire (Institute of Electrical and Electronics Engineers Inc., 2021-07-26)
    The security vulnerabilities of Electronic Health Record (EHR) systems are unique due to their interoperable nature and their specific methods of use and storage. A critical review was undertaken of existing and generalized maturity models, addressing cyber and information security scenarios in the healthcare sector. The study investigated how hospitals address their EHR system security via a survey that was distributed to private hospitals in Amman, Jordan. As a result of the study of maturity models that target cybersecurity, information security, healthcare, and big data, it was found that maturity models in the healthcare sector do not provide tools for determining maturity for characteristics and processes specific to hospitals. A Design Science Research Methodology was adopted as a reliable and robust mechanism to develop an EHR maturity model (EHR-MM) for hospitals, to assess the security of the EHR specifically and effectively in their organizations. A case study was used for evaluation, employing a template that addresses the effectiveness of the maturity models through domain expert reviews, providing positive results for the proposed approach.
  • Flight delay prediction: data analysis and model development

    Anees, Azib; Huang, Wei; University of Bedfordshire (Institute of Electrical and Electronics Engineers Inc., 2021-11-15)
    Flight delays in air transportation are a major concern that has adverse effects on the economy, the passengers, and the aviation industry. This matter critically requires an accurate estimation for future flight delays that can be implemented to improve airport operations and customer satisfaction. Having said that, a massive volume of data and an extreme number of parameters have restricted the way to build an accurate model. Many existing flight delay prediction methods are based on small samples and/or are complex to interpret with little or no opportunity for machine learning deployment. This paper develops a prediction model by analysing the data of domestic flights within the United States of America (USA). The proposed model gains insight into factors causing flight delays, cancellations and the relationship between departure and arrival delay using exploratory data analysis. In addition, Random Forest (RF) algorithm is used to train and test the big dataset to help the model development. A web application has also been developed to implement the model and the testing results are presented with the limitation discussed.
  • A 60 GHz broadband wearable antenna for body-to-body communications

    Zhang, Yutong; Yin, Rui; Shen, Xiaoliang; Yan, Na; Safdar, Ghazanfar Ali; Ur-Rehman, Masood; Fudan University; National Integrated Circuit Innovation Center; University of Bedfordshire; University of Glasgow (IEEE Computer Society, 2021-12-01)
    This paper presents a broadband antenna designed on flexible liquid crystal polymer (LCP) substrate for body-to-body short-range communications in 60 GHz frequency band. The antenna provides a wideband characteristic that covers the 51.1-70 GHz band centered at 60 GHz, owing to a coplanar waveguide (CPW) fed slotted patch structure. It attains a gain of 7.4 dBi and efficiency of 98% in the free space while 10.5 dBi and 67% in the on-body configuration. Through simulation and numerical analysis, paired antennas can cover at least 3.9-meter line-of-sight (LOS) body-to-body communications.
  • Flexible ultra-wideband antenna for 5G and beyond wearable applications

    Lyu, Bowen; Safdar, Ghazanfar Ali; Jamshed, Muhammad Ali; Ur-Rehman, Masood; University of Glasgow; University of Bedfordshire (IEEE, 2022-01-19)
    Millimeter wave frequencies are front running contenders for 5G wireless communications. A wideband antenna that can provide coverage in whole of these frequencies is a well sought off challenge. This paper presents design of a quasi-Yagi antenna consisting of three driven arms and three pairs of spiral directors that efficiently meets this requirement. To satisfy the flexibility demand of wearable applications, the antenna employs liquid crystal polymer LCP as the substrate. The antenna has a small size of 5×8×0.202 mm 3 while covering 24-71 GHz band.The proposed antenna achieves peak gain of 2.85, 6.33 and 8.52 dBi, respectively, and an efficiency of more than 70% in the desired frequency bands. The symmetric structure of the antenna also makes the fabrication easier. The ultra-wideband, radiation characteristics and flexibility of usage makes it a promising candidate for 5G and Beyond wearable applications.
  • Automotive security and theft prevention systems: state of the art

    Migacz, Lukasz; Feng, Xiaohua; Conrad, Marc; University of Bedfordshire (IEEE, 2022-03-15)
    This paper analyses automotive security and theft prevention systems available on the markets worldwide from 1949 to the present based on the literature available from the Internet. This analysis is motivated by the low usefulness of the security systems and an increasing number of car thefts. The goal of the analysis is to identify further research projects that should be considered while developing a system reliable enough to stop thieves and protect our cars. Moreover, the paper describes the origins of the Remote Keyless Entry and Remote Passive Entry systems, their function and their week sides by presenting examples of successful attacks performed against them. This paper also contains a proposition of improvements for the described systems. Furthermore, this paper is concluded with a summary of the information contained in this paper with a proposal of further work.
  • Artificial intelligence and blockchain for future cyber security application

    Feng, Xiaohua; Conrad, Marc; Eze, Elias; Hussein, Khalid; University of Bedfordshire (IEEE, 2022-03-15)
    AI (Artificial intelligence) application on Big Data had been developed fast. AI cyber security defense for the facing threats were required. Blockchain technology was invented in 2008 with BTC (Bit coin. This technology could be benefited alongside the custom of Blockchain, AI, Big Data and so on. There were a rapid progress in the advancement of Blockchain. This subject had recently become a discussion topic in the ICT (Information and Communications Technology) world. In this paper, AI security is discussed from the initial stage. Suggestion: In this paper, we discussed the impact of AI security from the initial stage and its impact and benefits to IT engineers, ICT students and CS (Computer Sciences) academic researchers, using a case study of medical records with personal recognizable identification privacy information that needs strict access control security. We considered its need for trustworthy cyber security, anti-fake, anti-alteration and transaction accounting transparency reputation to be applied to the NHS (National Health Service). Lastly, the paper provided some necessarily analysis. Blockchain technology had trustworthy cyber security, anti-fake, anti-alteration and transaction accounting transparency reputation to be considered to be applied to NHS (National Health Service). This short paper provided some analysis necessarily.
  • Robot, edge intelligence and data survey

    Feng, Ruomu; Feng, Xiaohua; Hebei University of Science & Technology; University of Bedfordshire (IEEE, 2022-03-15)
    A systematic approach review had been applied on the edge devices intelligence survey. The observing and recording events that occur in the process of the research had been carried out. This kind of events had been collected. The analysing of the case study data was carried out. Eventually, concluded some results based on the investigation. Furthermore proposed a solution for the kind of edge device data cyber security problem solving, Edge Intelligence was defined as the devices available at the edge layer have some limited amount of computing resources which can be utilized and incorporated with machine learning or AI algorithms to perform real-time data analytics [1]. Learning more in role of edge Intelligence to leverage IoT-assisted ecosystem. Here we also introduced Data Security Law 2021[2] application examples on Edge Intelligence. Data Security Law 2021 motivated from the cloud computing to edge device intelligence data process. This survey on edge intelligence had been carried out on investigated recent Robot and IoT (Internet of Things), AV (autonomous vehicle) and so on as edge terminal - edge devices and the challenges they had facing. In particular, AI (Artificial intelligence) application in edge device intelligence caused big data security threat. AI facing more challenges in edge computing. A comparison had been performed. Some advantages and disadvantages had been analysed. Edge device challenges on big data security issues had been discussed.
  • Anti-tailgating solution using biometric authentication, motion sensors and image recognition

    Akati, Jessica; Conrad, Marc; University of Bedfordshire (Institute of Electrical and Electronics Engineers Inc., 2022-03-15)
    Tailgating is a social engineering attack challenging physical security within organizations. It gained public traction in the year 1999 and has since remained a major concern in the field of security leading to the development of several anti-tailgating solutions. These solutions began with simple mechanisms like mechanical turnstiles, revolving doors, and man-trap systems and evolved into more modern technologies using infrared beams, 3D machine vision, face detection, BMI and face recognition combination, and an embedded solution using IP camera and video analytics. A critical analysis of these solutions uncovered certain weaknesses which run through most of them. These are the inability to detect two people side by side and the incapability of detecting multiple entries after a single access authorization. These shortfalls led to the development of the solution in this paper which aims to eliminate the shortcomings of existing technologies and boost security, by using a three-step anti-tailgating solution. The design science research methodology and aspects of qualitative and quantitative research are employed in designing a three-step anti-tailgating solution that combines face detection, palm recognition, and motion sensors, to eliminate the loopholes of existing technologies. The results from experimentation indicated that the face detection tool could detect two faces present. The motion sensors were shown to be efficient in performing people counting and detection, to eliminate tailgating and discrepancies in the number of entries against the number of authorized personnel. Integrated with palm recognition the overall system will function effectively because the three technologies complement each other's shortfalls, therefore preventing tailgating. It is concluded that this system will be an improved and more effective anti-tailgating solution.
  • Interference mitigation in device-to-device communications

    Ur-Rehman, Masood; Safdar, Ghazanfar Ali; Chaudhry, Mohammad Asad Rehman; University of Glasgow; University of Bedfordshire; Soptimizer (Wiley Telecom, 2022-12-31)
    Interference Mitigation in Device-to-Device Communications delivers a thorough discussion of device-to-device (D2D) and machine-to-machine (M2M) communications as solutions to the proliferation of ever more data hungry devices being attached to wireless networks. The book explores the use of D2D and M2M technologies as a key enabling component of 5G networks. It brings together a multidisciplinary team of contributors in fields like wireless communications, signal processing, and antenna design. The distinguished editors have compiled a collection of resources that practically and accessibly address issues in the development, integration, and enhancement of D2D systems to create an interference-free network. This book explores the complications posed by the restriction of device form-factors and the co-location of several electronic components in a small space, as well as the proximity of legacy systems operating in similar frequency bands. Readers will also benefit from the inclusion of: A thorough introduction to device-to-device communication, including its history and development over the last decade, network architecture, standardization issues, and regulatory and licensing hurdles An exploration of interference mitigation in device-to-device communication underlaying LTE-A networks A rethinking of device-to-device interference mitigation, including discussions of the challenges posed by the proliferation of devices An analysis of user pairing for energy efficient device-to-device content dissemination Perfect for researchers, academics, and industry professionals working on 5G networks, Interference Mitigation in Device-to-Device Communications will also earn a place in the libraries of undergraduate, graduate, and PhD students conducting research into wireless communications and applications, as well as policy makers and communications industry regulators.
  • Introduction to D2D communications

    Safdar, Ghazanfar Ali; Ur-Rehman, Masood; Chaudhry, Mohammad Asad Rehman; University of Bedfordshire; University of Glasgow; Soptimizer (Wiley Telecom, 2022-12-31)
    Device-to-device (D2D) communication enables researchers to merge together the achievements of long-term development in previously two disjoint networking techniques, that is, ad-hoc networking and centralized networking. D2D can support local area services such as content distribution, local advertisement, and location aware services very effectively through unicast, groupcast, and broadcast transmission. The diversity of cellular communication provides lots of flexibility for D2D communication in terms of link establishment, resource allocation, energy efficiency, as well as applications and services. D2D communication in cellular spectrum enables direct communication between devices that are in close proximity, and it is an exciting and innovative feature for next generation of cellular networks. In terms of spectrum usage, D2D is primarily classified into two categories, in-band and out-band. D2D provides benefit in terms of performance improvement, spectral, and energy efficiency, but there are few challenges while integrating D2D in long term evolution-advanced network.
  • Connected vehicles and motor factories of the future adopting 5G technology for vehicle-to-factory communications

    Rogers, Samuel Lear; Safdar, Ghazanfar Ali; Kalsoom, Tahera; Ur-Rehman, Masood; BMW; University of Bedfordshire; University of West of Scotland; University of Glasgow (Institute of Electrical and Electronics Engineers Inc., 2022-08-25)
    Since the inception of Henry Ford's 'production line' in 1913, continuous efforts have been made to enhance the efficiency of motor manufacture. Recently, a shift to Industry 4.0 has resulted in the digitisation of traditional production systems, by integration of specialist software applications. Concurrently, there have been efforts to develop and produce connected vehicles, with aspirations of autonomous features in the distant future. Achieving these concepts relies on the implementation of wireless communication media, such as 5G Vehicle-to-Everything (V2X) which is interoperable in the shop floor environment, and in the public realm. We review the key use-cases for 5G V2X in the context of motor factories, and how they can be used with private 5G networks, which leads to the development of the term 'Vehicle-to-Factory' (V2F). Based on this information, we explore the feasibility of a vehicle transitioning from a private 5G network to the public 5G network after its manufacture. Furthermore, we proceed to identify challenges which arise from the implementation of 5G, and explore novel approaches to overcome these, such as infrastructure sharing which uses existing street furniture to provide 5G coverage.
  • Breast cancer survival analysis with molecular subtypes: an initial step

    Zhang, Lingli; Wu, Jiaiun; Zhao, Youbing; Hu, Wenxian; Qin, Aihong; Dong, Feng; Liu, Enjie; Zeng, Hao; Xie, Hao; Du, Hui; et al. (Institute of Electrical and Electronics Engineers Inc., 2022-12-14)
    As a predominant threat to women’s health worldwide, breast cancer has become increasingly important in oncology research. The discovery of molecular subtypes of breast cancer has led to more subtype oriented treatment and prognosis prediction. Effective prognosis models help to estimate the recurrence as well as the quality and duration of survival, leading to more personalized treatments. However, most traditional prognostic models either ignore molecular subtypes or only make limited use of them. The roles of molecular subtypes in the development and treatment of breast cancer have not been fully revealed. With the over 1200 cases collected by Sir Run Run Shaw Hospital of Zhejiang University in the past two decades, we aim to improve understanding of molecular subtypes and their impacts on the prognosis via data analysis in the long run. As the initial stage, this short paper presents our preliminary work of logistic regression experiments with the data. Though molecular subtypes have not been included the tentative model, they are to be explored further in following stages.
  • Photorealistic true-dimensional visualization of remote panoramic views for VR headsets

    Livatino, Salvatore; Regalbuto, Alessio; Morana, Giuseppe; Signorello, Giovanni; Gallo, Giovanni; Torrisi, Alessandro; Padula, Gianluca; Pelc, Katarzyna; Malizia, Alessio; Farinella, Giovanni Maria; et al. (Institute of Electrical and Electronics Engineers Inc., 2023-06-14)
    Virtual Reality headsets have evolved to include unprecedented display quality. Meantime, they have become light-weight, wireless and low-cost, which has opened to new applications and a much wider audience. Photo-based omnidirectional imaging has also developed, becoming directly exploitable for VR, with their combination proven suitable for: remote visits and realistic scene reconstruction, operator's training and control panels, surveillance and e-tourism. There is however a limited amount of scientific work assessing VR experience and user's performance in photo-based environment representations. This paper focuses on assessing the effect of photographic realism in VR when observing real places through a VR headset, for two different pixel-densities of the display, environment types and familiarity levels. Our comparison relies on the observation of static three-dimensional and omnidirectional photorealistic views of environments. The aim is to gain an insight about how photographic texture can affect perceived realness, sense of presence and provoked emotions, as well as perception of image-lighting and actual space dimension (true-dimension). Two user studies are conducted based on subjective rating and measurements given by users to a number of display and human factors. The display pixel-density affected the perceived image-lighting and prevailed over better lighting specs. The environment illumination and distance to objects generally played a stronger role than display. The environment affected the perceived image-lighting, spatial presence, depth impression and specific emotions. Distances to a set of objects were generally accurately estimated. Place familiarity enhanced perceived realism and presence. They confirmed some previous studies, but also introduced new elements.
  • Swarm robot SLAM with point-line matching in weak information environments

    Guo, Jikai; Li, Dayou; Qiu, Renxi; University of Bedfordshire (Institute of Electrical and Electronics Engineers Inc., 2023-11-07)
    Simultaneous Localization and Mapping (SLAM) in weak information environment scenarios where there are only sparse and ambiguous references are available presents significant challenges. This paper introduces a swarm robots based point-line matching based solution to enhance the accuracy and robustness of localization and mapping within such environments. Our approach employs a centralized serverbased architecture for the swarm robot system and utilizes an ORB-based Distributed Bag-of-Words (DBoW) algorithm to construct point and line feature dictionaries. By generating point-line pairs, the proposed method effectively addresses perceptual overlap issues, thereby improving the overall performance of localization and mapping for swarm robots. Experimental results validate the superiority of the proposed approach in comparison to traditional methods, highlighting its potential for real-world applications in challenging environments for swarm robot SLAM.
  • Enhancing sparse data performance in e-commerce dynamic pricing with reinforcement learning and pre-trained learning

    Liu, Yuchen; Man, Ka Lok; Li, Gangmin; Payne, Terry R.; Yue, Yong; Xi'an Jiaotong-Liverpool University; University of Bedfordshire; University of Liverpool (Institute of Electrical and Electronics Engineers Inc., 2023-09-25)
    This paper introduces a reinforcement learning-based framework designed to tackle dynamic pricing challenges in e-commerce. Prior research has predominantly concentrated on algorithm selection to enhance performance in dense data scenarios. However, many of these models fail to robustly address sparse data structures, such as low-traffic products, leading to the 'cold-start' problem [4]. Through numerical analysis, our framework offers innovative insights derived from the design of the reward function and integrates product clustering with pre-trained learning to mitigate this issue. As a result of this optimization, the performance of predictive models on sparse data is expected to see substantial improvement.
  • VehA & PedA mobility based scheduling in future communication networks

    Ashfaq, Khuram; Safdar, Ghazanfar Ali; Ur-Rehman, Masood; University of Bedfordshire; University of Glasgow (Institute of Electrical and Electronics Engineers Inc., 2023-08-14)
    The primary means of communication is quickly evolving to be wireless communications. However, when compared to wired links, their properties render traffic susceptible to time and location-dependent signal attenuation, noise, fading, and interference, resulting in time variable channel capacity and link error rate. Scheduling algorithms are critical in wireless connections for ensuring quality of service (QoS) metrics such as throughput, latency, jitter, fairness, and packet loss rate. The scheduler is critical in current and future cellular communications because it assigns resource blocks (RB) to different users for transmission. The scheduling algorithm decides based on the link state, number of sessions, reserved rates, and the status of the session queues. The information required by a scheduler implemented in the base station can be easily gathered from the downlink stream. It evaluates the performance of four well-known scheduling methods, including round robin (RR), best channel quality indicator (BCQI), proportional fair (PF), and fractional frequency reuse (FFR), for dynamic use Pedestrian (PedA) and Vehicular (VehA). The performance of these four algorithms is measured in terms of throughput, fairness index, spectral efficiency, and overall effectiveness. System-level simulations were carried out utilizing a MATLAB-based LTE-A Vienna simulator.
  • Smart Grid 2.0 security

    Porambage, Pawani; Liyanage, Madhusanka; Yapa, Charithri; De Alwis, Chamitha; University of Oulu; University College of Dublin; University of Sri Jayewardenepura; University of Bedfordshire (IEEE, 2023-12-31)
    Smart Grid 2.0 is the next generation of Smart Grid technology that leverages advanced communication and automation technologies to improve the efficiency and reliability of energy systems. This chapter provides a comprehensive overview of the security aspect of Smart Grid 2.0. By the end of this chapter, readers will understand the taxonomy of Smart Grid 2.0 and the security attacks that are associated with it. They will also gain an understanding of the privacy objectives of Smart Grid 2.0 and the standardization initiatives that are being developed to ensure its security and privacy. With this knowledge, readers will be able to assess the security and privacy implications of Smart Grid 2.0 and make informed decisions about its implementation.
  • A systematic approach to the model development of reactors and reforming furnaces with fuzziness and optimization of operating modes

    Orazbayev, Batyr; Zhumadillayeva, Ainur; Kabibullin, Madyar; Crabbe, M. James C.; Orazbayeva, Kulman; Yue, Xiao-Guang; L. N. Gumilyov Eurasian National University; Oxford University; University of Bedfordshire; Esil University; et al. (Institute of Electrical and Electronics Engineers Inc., 2023-07-12)
    The paper studies the problems of developing interconnected models of aggregates of complex chemical-technological systems (CTS) in conditions of scarcity and fuzziness of the initial information. Since the known methods of model development do not allow for solving these problems, we propose a systematic method that allows the development of a package of models of interconnected aggregates of complex CTS using the available information of a different nature. In the proposed method, formal and informal methods of system analysis are jointly used. Due to synergy and emergence, the proposed system allows the synthesis of the most adequate and effective CTS models using the knowledge, experience, and intuition of experts and other available data. Using the method, we have developed a package of models of interconnected reactors and used it in a case study for a reforming reactor and furnace of the Atyrau refinery catalytic reformer based on available statistical and fuzzy information. To study and optimize the operating modes of the reformer unit, a system of computer simulation and optimization of the reformer unit was created. Comparison of the obtained results of computer modeling and optimization with the results of known deterministic models shows the advantages of the proposed approach in the face of scarcity and fuzziness of the initial information. The importance and novelty of the proposed method lie in the possibility of developing a package of models of interrelated CTS units based on various types of available information. This allows one to systematically simulate and optimize the operating modes of the CTS. The practical significance of the results is that this case study can be successfully applied in the development of models of various technological units for oil refining, petrochemistry, and other industries.

View more