10th International Congress on Information and Communication Technology in concurrent with ICT Excellence Awards (ICICT 2025) will be held at London, United Kingdom | February 18 - 21 2025.
Sign up or log in to bookmark your favorites and sync them to your phone or calendar.
Authors - Rotimi Williams Bello, Pius A. Owolawi, Chunling Tu, Etienne A. van Wyk Abstract - One of the mainstream methods for user identification has been by face. However, the vulnerability of face-swapping applications to security issues when swapping the faces between two different facial images, has undermined the genuine aims of the technology, thereby threatening the security of certain applications and individual users when such action is performed without caution. To address this, we propose the development of scalable and safe cloud architecture for a face-swapping application that lets users upload two photos and get a face-swapped output. This is achieved by: (1) creating a secure virtual private cloud (VPC) to hold all application resources, (2) using a Web Application Firewall (WAF) to filter and safeguard requests, (3) putting application programming interface (API) Gateway into place to provide regulated access to the application's API, (4) processing and overseeing face-swapping operations with Lambda functions, (5) using VPC Endpoint to store input and output photos in Simple Storage Service (S3) buckets for private access, and (6) configuring a Simple Notification Service (SNS) to inform users of the progress and completion of their requests. A face swapping dataset derived from an open benchmark dataset was utilized for training and testing the proposed system. The experiment produced an effective solution with a 93% detection accuracy. The implications of this solution are: (1) the provision of security and private access to Amazon Web Services (AWS) by VPC Endpoints and WAF, (2) elimination of Network Address Translation (NAT) Gateway costs by utilizing VPC Endpoints for private S3 access, (3) offering of a scalable processing environment by Lambda functions without the need for server management, (4) delivering of real-time notifications by SNS to users regarding their request status, and (5) optimization of S3 storage ensures quick and efficient access to images.
Authors - Duong Bui, Cuong Nguyen Abstract - Purpose - We are currently experiencing the era of digital transformation. This leads to concepts like digital leadership, continuous learning, and digital culture. The objective of this study is to investigate the influence of various dimensions of digital leadership (DL) on enhancing responsible innovation (RI), with a mediating role of continuous learning (CL) and the management of digital culture (DC). Design/methodology/approach – This study was employed a mixed-methods research design. Data were collected using a self-administered questionnaire distributed to a sample of 250 employees from small and medium-sized enterprises in Vietnam, selected through convenience sampling. Findings - Structural equation modeling was utilized for path analysis in the study. The findings indicated a positive and significant impact of digital leadership (DL) on responsible innovation (RI), mediated through the roles of continuous learning (CL) and digital culture (DC). Practical implications – This study has highlighted the significance of the impact of DL to create CL and DC in Vietnam. The study also confirmed the relationship between DL and RI. It adds to the evidence on digital leadership in Vietnam. Originality/value – Empirical evidence was provided by this study to support the role of DC in fostering RI. Furthermore, how DL strengthens its influence on CL and DC within organizations was demonstrated. By doing so, a critical gap in understanding the impact of DL on RI, CL, and DC in the context of Vietnam is addressed by this research.
Authors - Suresh Rasappan, S. Ahamed Nishath, Francis Saviour Devaraj Abstract - This study proposes a hybrid CNN-ANN architecture for lung cancer image classification on the LC105K dataset. Enhanced feature extraction and representation techniques improve classification accuracy. The model leverages CNN and ANN strengths, demonstrating superior performance compared to existing methods. Results show significant accuracy, precision, and recall improvements, offering a promising solution for computer-aided diagnosis.
Authors - Abdellah Tahenni, Abdelkader Belkhir Abstract - This study provides the privacy concerns of AI predictive algorithms for E-health systems. A significant disadvantage is that these algorithms can infer delicate private health data of people, particularly high profile figures, from big datasets. This might be infringing privacy and result in discrimination or safety threats. The paper additionally analyzes the danger of AI prediction algorithms escalating wider privacy violation risks for patients and providers like accidental disclosure of personal details or unauthorized use of system vulnerabilities for information theft via AI models. The mixed-methods methodology encompasses evaluation of AI algorithm abilities, privacy breach case studies, expert interviews, healthcare provider surveys and eHealth method penetration tests. The results plot vulnerabilities, risk levels and technical, cultural and regulatory variables related to these privacy risks. To lessen those risks, a framework is suggested that has specialized safeguards including AI auditing and differing privacy, governance (data security policies and ethical AI guidelines), organizational (devoted privacy roles and staff training) along with ethical considerations balance innovation with privacy protection. Lastly, the study suggests multi-stakeholder, strategic and collaborative interaction among healthcare, policymakers, AI designers and patient advocates to mitigate AI driven privacy issues in eHealth systems through serious scrutiny and suggestions guided by this vision.
Authors - Kevin Tanuwijaya, Elfindah Princes Abstract - Indonesia's digital economy is rapidly expanding, fueled by technological advancements and government support for e-commerce. E-commerce has become a cornerstone of the nation's economy, significantly contributing to Gross Merchandise Value (GMV). However, Small and Medium-Sized Enterprises (SMSEs), crucial to Indonesia's economic landscape, face challenges in building lasting customer loyalty. This study investigates the impact of affiliate marketing and gamification on SMSE sustainability, focusing on economic, social, and environmental dimensions. Data were collected from a purposive sample of 100 respondents in Jakarta was analyzed using Structural Equation Modeling with Partial Least Squares (SEM-PLS). The results demonstrate that both affiliate marketing and gamification directly enhance customer loyalty. Furthermore, customer loyalty was found to have a significant positive impact on SMSE sustainability. Crucially, the study reveals a mediating effect of customer loyalty, bridging the gap between affiliate marketing and gamification strategies and their ultimate impact on business sustainability. This research contributes valuable insights into the sustainable business literature by empirically examining its effects on key sustainability variables. The study concludes with a discussion of theoretical implications, practical recommendations for SMSEs, and avenues for future research.
Authors - Nguyen Quoc Cuong, Huynh Gia Nghi, Mai Thi Bich Ngoc Abstract - Live streaming has transformed online purchasing, especially for tech-savvy Generation Z, who remain cautious in their purchasing decisions. This study explores the motivational factors driving live streaming-enabled purchasing decision in Vietnam. Applying CB-SEM in the Theory of Planned Behavior, this papers analyzes relationships among 07 variables: information quality, streamer attractiveness, interaction quality, trustworthiness, streamer expertise, online purchase intention, and online purchase decision. Samples consist of 233 Gen Z residents in Ho Chi Minh City (April–June 2024) was analyzed using SPSS and AMOS.. The findings indicate that all examined variables positively impact Gen Z's live stream purchase decisions, helping to advance scholarly understanding and offering insights for businesses to effectively integrate live streaming into their omnichannel strategies.
Authors - Ansh Soni, Krish Modi, Aneri Shah, Nishant Doshi Abstract - A successful and efficient transportation system depends on the credibility of engines and machinery. With the help of NASA Turbofan Jet Engine dataset, this paper focuses on the predictive maintenance framework to boost transport efficiency by leveraging sensor data. With the help of machine learning algorithms, we predict the Remaining Useful Life (RUL) of engine components based on training the model with appropriate algorithms that prompt scheduled services and maintenance to reduce downtime. Feature engineering techniques and predictions of RUL, Health Index(HI), and degradation score- the proposed model provides a methodology for enhancing system dependability and minimizing maintenance costs. This study provides valuable insights into current transportation setbacks.
Authors - Dario Lombardi Abstract - This study investigates the impact of Serious Games (SGs) and ChatGPT on nursing education, focusing on usability, learning outcomes, engagement, skill transfer, and future usage intentions. Using a multi-phase design aligned with the ADDIE instructional model, the study explores how these tools facilitate learning through experiential and cognitive pathways. In the first phase, students designed instructional interventions using the ADDIE model, while the second phase introduced ChatGPT as an AI-driven support tool. Findings reveal that SGs promote experiential learning by enhancing clinical skills, emergency response, and procedure retention. Usability metrics were high, with 79% of participants rating interface intuitiveness positively. Conversely, ChatGPT supported cognitive scaffolding, enabling faster and more effective instructional design. Students reported a significant increase in familiarity with AI, with 68.4% moving from "low" to "medium" familiarity. Engagement and motivation were strong for both tools, with 84.2% of participants intending to continue using ChatGPT. Despite its small sample size (n=19), this study highlights the potential for hybrid models that integrate SGs and AI to improve nursing education. It calls for the incorporation of these tools into curricula, emphasizing experiential learning (SGs) and cognitive support (AI), thereby addressing both procedural and conceptual learning needs.
Authors - Ravichandran S, R Kasturi Rangan, Manjesh R, S Karthik, T N Hemanth Abstract - The page-level recognition and reordering of handwritten documents is important in digitizing and archiving systems. These systems focuses on solving the two problems of converting relevant parts of a handwriting document into recognizable formats for machines as well as correctly sequencing pages in order to preserve context. Building upon the state-of-the-art in Optical Character Recognition (OCR) and Document Layout Analysis (DLA), The paper suggests that these methods are effective for page-level text recognition that combines automatic reading order detection with advanced OCR modeling. This study evaluates the impact of a hybrid architecture combining Vision Transformers (ViT) for powerful feature extraction, and transformer-based Language Models (LMs) to provide context during text decoding. We then pose the task of reordering as a sorting problem and use a pairwise order-relation operator trained from annotated data to generalize to various layouts of input documents. The phenomenon under study reveals significant trends in the state-of-the-art performance on standard datasets with significant recognition accuracy gain and reordering precision. It opens up the efficient processing of handwritten documents in applications that range from preserving historical writing samples to today’s administrational scanned handwritten documents.
Authors - Khanyisani I Ndlovu, Timothy T Adeliyi, Alveen Singh Abstract - Integrating Information and Communication Technology (ICT) in educational settings has become fundamental to modern pedagogy. Despite its significant contributions, ICT's adoption, integration, and usability pose challenges for school students in developing countries, particularly in Sub-Saharan Africa. This study aims to identify and analyse the diverse factors influencing the widespread adoption of ICT among students. A comprehensive systematic literature review revealed thirty-eight factors from eighty-four articles that either facilitate or hinder ICT integration in students' academic activities. Using Principal Component Analysis (PCA) a statistical method known for reducing dimensionality and uncovering patterns in complex datasets. The study extracts the most critical factors impacting ICT adoption. The findings indicate that in addition to over-digitalization, cognitive barriers, health issues, time constraints, funding limitations, and a lack of modern software are significant factors affecting students' engagement with technology. The implications of this study are relevant to policymakers, educators, and academic institutions, providing a data-driven basis for strategies aimed at improving ICT adoption
Authors - Pesta Evaria Simbolon, Himsar Silaban, Khasan Effendi, T. Herry Racmatsyah Abstract - Currently, the professionalism of expert staff at the House of Representatives (DPR RI) is influenced by various interconnected factors such as work management, coordination, compensation, qualifications, and competence. This study aims to identify the key factors affecting the professionalism of expert staff at DPR RI and provide recommendations to improve their performance. A qualitative research method was employed using thematic analysis, grounded theory, and triangulation through NVIVO 14 to analyze interview data and relevant documents. The results indicate that work management and coordination, competitive compensation, staff qualifications and competencies, rigorous recruitment and selection processes, and objective performance evaluations are the primary factors influencing professionalism. Furthermore, professional development and clear career paths enhance staff loyalty and performance, while benchmarking against industry standards offers insights into best practices that can be adopted. Addressing challenges and barriers in the workplace and ensuring a clear organizational structure help define roles and responsibilities. High-quality support from expert staff and satisfaction with compensation has a direct impact on the effectiveness of DPR RI. Thus, DPR RI must focus on and improve these factors to ensure the optimal efficiency and effectiveness of expert staff, ultimately supporting the overall performance and effectiveness of DPR RI.
Authors - Ilnar Yarullin, Ramis Nasibullov, Almaz Galimov, Shamil Sheymardanov Abstract - This paper is important because it highlights the need to improve how we help students with their internships. This need arises from higher expectations for the quality of training future teachers receive and the limitations of traditional methods of internship support. The article explains the idea and functioning of a digital platform designed to assist students with their internships. The aim of this platform is to improve the quality of their professional training by enhancing communication among everyone involved in the educational process. The software described provides options for creating flexible schedules that consider different types of internships, such as in-person, remote, and hybrid. This flexibility is especially important when working with various groups of students. In addition to assigning mentors and supervisors, both teachers and students can outline the goals and objectives of the internship. This helps to avoid misunderstandings and ensures that everyone is clear about what needs to be accomplished at each stage of the internship. The platform is connected to the university's corporate system, teaching websites, and electronic publications. As a result, teachers, students, and administrators can share information more easily.
Authors - Vikas Shah, Travis Rice, Aarav Shah, Aarush Shah Abstract - Artificial intelligence (AI) empowered the transformation of decision-making processes for lending services, delivering improved efficiency, scalability, and precision. However, the adoption of AI in loan origination and application processing has introduced significant ethical challenges, including recognizing biases, fairness, transparency, reliability, and accountability. The paper identifies primary challenges in automated lending services (ALS), AI-enabled decision-making, and the deriving of AI governance practices. This paper proposes a deterministic framework (DF) designed to systematically identify and address the ethical dimensions of AI in lending services. The DF spans comprehensive mechanisms encompassing data collection, preprocessing, model development, deployment, monitoring, and governance. Core ethical dimensions of explainability, transparency, and equitable outcomes are recognized within the governance lifecycle stages. The DF continuously integrates novel industry regulatory standards and governance methodologies to identify, measure, and mitigate ethical risks, ensuring operational efficiency and adherence to ethical principles. This research provides a structured approach grounded in deterministic methods, enabling measurable, repeatable, and auditable business processes to enable trust and accountability in AI-driven ALS. An empirical case study focusing on ALS for students and their families is presented to evaluate the DF's applicability and effectiveness. The findings provide actionable insights for financial institutions, policymakers, and technologists seeking to implement ethical AI practices, strengthen risk management, and deliver equitable and accountable lending services to diverse populations.
Authors - Oleksandr Kurbatov, Lasha Antadze, Ameen Soleimani, Kyrylo Riabov, Artem Sdobnov Abstract - This article proposes an extension for privacy-preserving applications to introduce sanctions or prohibition lists. When initiating a particular action, the user can prove, in addition to the application logic, that they are not part of the sanctions lists (one or more) without compromising sensitive data. We will show how this solution can be integrated into applications, using the example of extending Freedom Tool (a voting solution based on biometric passports). We will also consider ways to manage these lists, versioning principles, configuring the filter data set, combining different lists, and using the described method in other privacy-preserving applications.
Authors - Asmaa Berdigh, Kenza Oufaska, Khalid El Yassini Abstract - The inclusion of wireless communication technologies in the car industry has drastically reshaped it over the last two decades. This evolution began with the integration of Bluetooth and Smart Key technology in 1998 and has since progressed to the application of wireless communication in the cell monitoring controller of battery management systems. The objective of this study is to conduct an SMS to evaluate the research investigating wireless communication technologies used in the in-vehicle network. We adopt the Systematic Mapping Study (SMS) approach, a rigorously defined research methodology with roots in the medical and software engineering, using defined criteria to filter out the research contributions stored in both Scopus and GoogleScholar databases over the last twenty years. We synthesize the resulting data and produce this article. This work aims to organize information from studies published within this disciplinary field over the past two decades, presenting them in a systematic map form, and discussing the results and their implications for future research, concluding by a visual display of which automotive domains have the largest wireless communication. The SMS provides a clear and comprehensive picture drawn from precise questions. The derived outcomes could hold significance for both researchers and industry professionals considering the integration of wireless communication technologies within in-vehicle networks. Despite the substantial body of research identified over the past 20 years on wireless technologies, only a limited number of studies have specifically focused on the In-Vehicle Networks.
Authors - Esther Endjala, Hanifa Abdullah, Mathias Mujinga Abstract - This paper explores the theoretical and strategic foundations for cultivating a cybersecurity culture within telecommunications institutions. Drawing on established behavioral theories Social Cognitive Theory (SCT), Protection Motivation Theory (PMT), Theory of Planned Behavior (TPB), and Technology Acceptance Model (TAM) it examines the opportunities for enhancing cybersecurity awareness and transforming employee behaviors into a resilient human firewall. The paper synthesizes existing literature to highlight the role of leadership, employee engagement, training, collaboration, and recognition in fostering a cybersecurity culture. The review further identifies gaps and limitations in the current approaches, proposing a conceptual foundation for developing an effective cybersecurity culture framework tailored to telecommunications institutions. Appropriate cybersecurity culture is essential in developing the entity and helps protect organizational assets such as data, networks, and systems when technical defenses are quite significant. The section takes into consideration the theoretical aspect of cybersecurity culture and comes out with a derived underlying framework that incorporates aspects like the Social Cognitive Theory, Protection Motivation Theory, Theory of Planned Behavior, and Technology Acceptance Model. Creating a strong cybersecurity culture faces several significant challenges, including resistance to change, limited resources, and regulatory hurdles. The proposed framework emphasizes the importance of top management commitment, employee involvement, ongoing training, and interdepartmental collaboration as vital components for cultivating this culture. Organizations should address these challenges to successfully establish an effective cybersecurity culture. By integrating these elements, cybersecurity can become a fundamental organizational value, enhancing awareness, compliance, and employee engagement. This paper establishes the groundwork for further research on cybersecurity culture and outlines key steps to strengthen organizational resilience and ensure a safe digital environment.
Authors - Amna Altaf, Zeashan Khan, Adeel Mehmood, Jamshed Iqbal Abstract - Advancements in Artificial intelligence (AI) and robotics call for prioritising research in education and pedagogy in these domains including teachers’ training to practice emerging learning and teaching strategies. This paper explores the model of teachers’ training by using CDIO (Conceiving, Designing, Implementing and Operation) framework as a guide and a robotic platform as an example in AI education. In the pilot study, nine workshop sessions were designed and organised for a group of five teachers introducing them to robotics-led delivery of AI content using a mobile robot ‘Duckiebot’. The prominent workshop contents include robot vision, object detection, state estimation and localisation, task planning and reinforcement learning. Preliminary results in the form of feedback from the workshop participants demonstrated that the teaching model presented in this study made a promising contribution in terms of improving teachers’ intellectual and pedagogical skills as well as their confidence in achieving learning outcomes. The presented CDI-based robotics-led AI teaching model adds to the dialogue on innovative AI and engineering education methods. It is anticipated that wider dissemination of the findings in this research will lay the groundwork for a larger educational impact.
Authors - Muhammad Sufyan Akbar, Guo Jiandong, Muhammad Irfan Khan, Asif Iqbal, Salim Abstract - This paper introduces a deep learning-based approach for point cloud classification, leveraging the PointNet architecture to optimize 3D object recognition. The method effectively addresses the challenges associated with unordered point cloud data, achieving superior classification performance with 92% accuracy, 91% precision and recall, 89% F1-score, and 96% sensitivity and specificity. The proposed model captures spatial features directly from raw point cloud data, demonstrating its potential for real-world applications in 3D object recognition and scene understanding. Comprehensive experiments on benchmark datasets validate the model’s effectiveness in classifying complex 3D structures, highlighting its robustness and efficiency. Future research will focus on advancing feature extraction techniques and refining the model to enhance classification performance under more demanding conditions.
Authors - Timothy T Adeliyi, Funmi Adebesin, Edidiong R Umoh Abstract - Unemployment remains a persistent challenge for both developed and developing countries, leading to the underutilisation of resources. Many Sub-Saharan countries experience high unemployment rates due to weak economic indicators. This study adopts a data-driven approach to investigate women's unemployment in Sub-Saharan Africa, with a focus on the factors contributing to employment disparities and advancing gender equality, for Sustainable Development Goal 5 (SDG 5). Using Multiple Correspondence Analysis (MCA), the research identifies and analyses key factors that contributes to the high unemployment rates among women in the region. The findings reveal significant links between unemployment and factors such as age, region, and wealth index. By shedding light on these disparities, the study offers a comprehensive understanding of the structural barriers faced by women in the labour market. The results emphasise the need for specific policies and interventions to combat gender inequality and boost women's economic participation to achieve SDG 5. This research enriches the broader dialogue on sustainable development and gender equality, providing crucial insights for policymakers and stakeholders working towards more inclusive labour markets in Sub-Saharan Africa.
Authors - Salimah Saeid, Tahani Almabruk, Muetaz Abdulsamad Abstract - The diagnosis of ovarian tumors remains a challenging task due to the inherent variability and complexity of imaging data. This study evaluates the efficacy of transfer learning and fine-tuning techniques in convolutional neural net-works to enhance the classification accuracy of ovarian tumors in ultrasound images. The performance of YOLOv8 and VGG16 models were compared, including a modified VGG16 architecture optimized for this application. YOLOv8 models were evaluated both from scratch and with pre-trained weights, while VGG16 was employed for feature extraction and fine-tuning. The Modified VGG16 outperformed all other models, achieving the highest classification accuracy (%90) and the shortest training time (8.63 hours). Advanced data augmentation strategies and architectural optimizations effectively addressed issues such as class imbalance and overfitting. These results highlight the potential of customized CNN architectures and transfer learning to improve diagnostic accuracy and efficiency, advancing the development of reliable tools for ovarian tumor classification in clinical imaging.
Authors - Mohammad Anwar Rahman, Rafiul Hassan Abstract - Accurate prediction of lithiumion batteries' state of health (SOH) is crucial for preventing catastrophic system failures. This study investigates the application of ensemble modeling to characterize capacity degradation and fore-cast remaining charge-discharge cycles. Leveraging NASA's battery charge/discharge dataset, we developed and compared feed-forward neural network (FNN) and random forest (RF) regression models. To enhance predictive accuracy, we constructed an ensemble model that combines the strengths of both individual models. A key aspect of our methodology was the accurate evaluation of model performance across different battery datasets. Rather than using a single dataset for training and testing, we adopted a cross-validation approach to assess model generalization capabilities. This strategy allowed us to identify the robustness of the models for predicting SOH and estimating remaining battery life. Our findings indicate comparable performance among the FNN, RF, and ensemble models. While all models demonstrated effective capacity degradation prediction, the ensemble model exhibited slightly superior performance in a few scenarios. These findings emphasize the advantages of ensemble modeling in enhancing the accuracy and reliability of lithiumion battery prognostics.
Thursday February 20, 2025 4:15pm - 5:45pm GMT
Virtual Room DLondon, United Kingdom
Authors - Dattatraya Adane, Lakshya Agrawal, Shradha Wangota, Sharvari Inamdar Abstract - In order to overcome issues with traditional voting, such as voter impersonation, booth capturing, and logistical inefficiencies, this study proposes a biometric voting method. The technology guarantees that only authorised individuals can cast ballots while maintaining anonymity by combining fingerprint authentication with secure digital platforms. It features a web-based interface for election managers to manage candidates, track votes, and display real-time results, as well as a mobile app for voter registration and remote voting, improving accessibility. The solution lowers costs while increasing security, transparency, and engagement by utilising technologies like Firebase Firestore, the Mantra MFS100 fingerprint scanner, and QR code integration. This essay examines its design, use, and effects on modernising elections to promote efficiency, inclusivity, and trust.
