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 - Koichi Akashi, Hibiki Ito, Atsuko Yamashita, Katsuhiko Murakami, Sayaka Matsumoto, Kunihiko Takamatsu, Tetsuhiro Gozu Abstract - In recent times, a ground-breaking approach termed abridgement has been advocated for assessing reading comprehension of students in a distinct manner. The operation requires them to shorten the size of the given passage by only deleting the words or characters in it, while not permitting adding, paraphrasing, or swapping any of them. Thanks to its simple operational property, an efficient computational marking scheme has been invented employing the technique of dynamic programming (DP) by the following studies. These endeavors enabled further analyses to proceed, allowing for an ensuing research of reflections about teachings of abridgement collected from students using co-occurrence networks, to give an instance. Furthermore, evolution of this educational approach has led the researchers to suggest its unprecedented application to detect the presence of students who have potential difficulty with learning or specific tasks at an early stage, which is expected to contribute to the reduction of dropout rate in the long run. To make this happen, this research adopted the idea of utilizing large language models (LLMs), specifically Bidirectional Encoder Representations from Transformers (BERT) developed by researchers at Google in 2018, to automatically quantify the confidence levels of their understanding based on the technique of sentiment analysis of collected reflections. Combining grades calculated from DP-based algorithm with LLM newly invited to the research of abridgement, certain criteria have been established to issue warnings to identify those who are experiencing challenges early on.
Authors - Joao Bachiega Jr, Breno Costa, Leonardo R. Carvalho, Aleteia Araujo, Rajkumar Buyya Abstract - The fog computing paradigm allows for the distribution of computing resources and services at the edge of the network, close to end users, complementing cloud computing. Due to the dynamicity of fog computing environments, resource discovery is a key process that aims to find new computational resources that are available to integrate into it. These resources compose fog nodes, devices considering computational capabilities (such as CPU, memory, and disk) and behavioral characteristics (such as availability, scalability, and mobility). Performing an optimized resource discovery with all those attributes is still a challenge. This article proposes an efficient approach to resource discovery in fog computing that considers the computational capability and behavioral characteristics to select fog nodes. The results show that it is at least 33% more efficient than a similar solution found in the literature.
Authors - Kostadin Nevrokopliev, Silvia Gaftandzieva, Svetoslav Enkov Abstract - Every organization needs business intelligence reports and analytics to help it make informed decisions. This creates the need to develop a framework for composing business intelligence reports and analytics. This article describes the structure of a software framework that allows reports to be grouped by given fields and the data to be aggregated. The reports are composed in a way that allows a broader view of the data, which helps managers in their decision-making process. Its flexible and versatile structure allows reports to be customized to fit the specific needs of a given organization. The paper discusses sample reports developed using the framework and shows how they help higher education institutions to manage their activities effectively.
Authors - Quang-Vinh Dang Abstract - Product retrieval in e-commerce systems has traditionally relied on text-based matching between user queries and product descriptions. While recent advances have introduced image-based search capabilities leveraging deep learning techniques, existing systems typically operate in isolation, processing either textual or visual queries independently. However, contemporary user behavior increasingly demonstrates the need for multi-modal search capabilities, particularly as smartphones enable users to seamlessly combine photographic content with textual descriptions in their product searches. This paper presents a novel multimodal retrieval augmented generation (RAG) framework that unifies text and image inputs for enhanced product discovery. Our approach addresses the limitations of conventional single-modality systems by simultaneously processing and correlating both visual and textual features. By leveraging the complementary nature of these modalities, our system achieves more nuanced and contextually aware product matching. Experimental results demonstrate that our multi-modal RAG framework significantly improves search accuracy and relevance compared to traditional single-modality approaches. Furthermore, user studies indicate enhanced satisfaction and reduced search friction, suggesting meaningful improvements to the e-commerce user experience. Our findings contribute to the growing body of research on multi-modal information retrieval and offer practical insights for implementing more sophisticated product search systems in commercial applications.
Authors - Nur Anneliza Abd Latip, Hanita Hanim Ismail, Harwati Hashim, Wardatul Akmam Din Abstract - Textbook role in English language education is imperative in providing curriculum structure, standardising instruction, and maintaining the quality of teaching. Often it is an efficient material for teachers as it is readily available and reliable. Nonetheless, the one-size-fits-all material may not reflect real life language outcome. In reading section of textbook, the passages are simplified, adapted and inauthentic. In a class geared towards high stake reading examination, textbook is needed as it is mirroring language test format, making it a mock test situation. This research exposed students in tackling reading class to texts from periodical literature and explored them using an online whiteboard. The research is not striving to replace textbook, rather put forwards materials that provide learners with opportunity to have an active role in directing their own cognitive resources in reading. The study aims to investigate whether the intervention of using these materials have an effect towards reading achievement. A quasi-experimental design was applied to see quantitative evidence of reading marks. The participants were pre-university students studying at Universiti Malaysia Sabah, one of the public institutions in Malaysia. The findings show that students who are using the periodical literature and online whiteboard revealed significant changes and higher marks of reading achievement compared to the group of students who are only using textbook. The study is hoped to supplement reading class with relevant materials of authentic text and current technology.
Authors - Makoto Hirano, Kayoko Yamamoto Abstract - The purposes of tourism have been becoming diversified in recent years, and a form of travel known as food tourism is becoming increasingly popular. However, little research has been conducted on systems that support food tourism. Against such a backdrop, the present study aims to design, develop, operate and evaluate a food tourism support system that is supported to decide on restaurants for lunch and dinner, tourism spots to visit along the way, and routes to visit these destinations. The system comprises an original tourism plan creation system, web geographic information systems (Web-GIS) and web-augmented realty (Web-AR). In the present study, a location-based Web-AR system is developed. The system was operated for 30 days from December 22, 2023 to January 20, 2024, in Central Yokohama City of Kanagawa Prefecture, Japan. Total number of users was 50 and 20 tourism plans were created during the operation period. Based on the evaluation results, it is clear that the principal functions and the overall system were highly evaluated, regardless of food tourism experience or advance creation of tourism plan. Furthermore, it is evident that there was a high number of visits to the pages for most of the principal functions, and the system was used in a manner consistent with the purpose of the present study.
Authors - Mithil Mistry, Hasti Vakani, Hardikkumar Jayswal, Nilesh Dubey, Mann Patel, Jai Mehtani, Dipika Damodar, Ayush kariya Abstract - This research explores the integration of advanced deep learning techniques for automated pothole and speed bump detection, highlighting various methodologies from recent literature. A systematic review reveals that models utilizing Convolutional Neural Networks (CNNs), including YOLOv3, VGG16, and EfficientNetB0, have achieved impressive accuracy rates in real-time road surface monitoring. In particular, the EfficientNetB0 model was finetuned using a comprehensive dataset comprising 400 annotated images of potholes and speed bumps, collected under diverse environmental conditions. The model achieved a validation accuracy of 91.91%, demonstrating robust performance in identifying road anomalies. Notably, the implementation of advanced data augmentation and regularization techniques, such as dropout and L2 regularization, contributed to preventing overfitting and enhancing generalization across varying contexts. This study underscores the potential of deep learning frameworks in improving road safety and maintenance efficiency, paving the way for future enhancements, including dataset expansion and real-time application integration.
Authors - Ishaan Bhattacharjee, Pranav H P, Harish Satheesh, Disha Jain, Bhaskarjyoti Das Abstract - Satirical content is notoriously difficult to detect, even humans often struggle to discern satire from genuine news. While significant strides have been made in computationally modeling textual satire using supervised learning, the challenge of detecting satire in multimodal content—combining both text and images—remains largely unexplored. In our research, we aim to address this gap by leveraging existing frameworks and tools to detect and differentiate multimodal satire from true news content. Satire builds on two key factors, i.e., knowledge and incongruity. Knowledge has two parts, i.e., local knowledge that is resident in image and text and global contextual knowledge that is not part of the content. Incongruity typically occurs between the first and second parts of the text. In this work, we present a three-step framework. First, we investigate multimodal frameworks such as BLIP, relying on its global knowledge without explicitly modeling the incongruity. Second, we attempt to model incongruity by focusing on the semantic gap between two parts of the text content while using a large language model in knowledge enhancement and next-sentence prediction. Finally, we combine the above two models utilizing local knowledge, global knowledge, and incongruity to offer class-leading performance. The investigations described in this work offer novel insights into the detection of satire in complex, multimodal content.
Authors - James Uys, Gunther Drevin, Lynette Drevin Abstract - Serious games have seen a rise in popularity as an alternative method to deliver information to learners. A problem that is often faced is maintaining learner engagement during the educational process. To address this, it is important to identify the elements which are essential to keeping a learner motivated during the learning experience. This research focused on motivational drivers behind learning as well as prevalent characteristics in serious games. These elements were then integrated into the development life cycle of a serious game. The developed artefact was evaluated using the RETAIN model.
Authors - Lloyd L.K. Modimogale, Jan H. Kroeze, Corne J. van Staden Abstract - The paper presents the results of a quantitative study examining the impact of the Fourth Industrial Revolution (4IR) on the South African coal mining sector, specifically focusing on reskilling. As mechanization and digital technologies increasingly permeate the industry, this study investigates the implications for employment and skill requirements among coal miners. Data on job displacement rates, skill sets, and reskilling initiatives within the coal mining workforce were collected using statistical analysis. The findings indicate a significant decline in demand for traditional low-skilled job roles, highlighting the urgent need for reskilling programs to facilitate workforce adaptation to new technological demands. The study aims to understand how coal miners view the impact of digitalization on the coal mine and the management of the reskilling process. The research highlights the need for proactive measures to mitigate job insecurity and ensure that workers remain relevant in a rapidly changing economic environment, thus contributing to the broader discourse on sustainable labor practices within the mining sector.
Authors - Mafalda Reis, Lidia Oliveira, Catarina Feio Abstract - The main objective of this research is to study the role of online influencers in the information consumption of young people in Portugal. It studies the information consumption habits of young people aged between 18 and 30, as well as their views on the role of influencers in transmitting information on social media and their relationship with the concept of opinion leadership. A literature review was conducted since 2018 to generate a view of the state of the art on the issue under study. The data collection instrument was a questionnaire survey, obtained using an interpretative methodology, through a quantitative analysis of the data. From the analysis of the 322 respondents to the questionnaire, it was concluded that: young people consume more information on social networks; women and students are the ones who follow influencers the most; the relationship between influencers and followers is not one of friendship; young people are mainly interested in health issues and national news; young people have already learned about a current issue because an influencer talked about it; young people consider influencers to be active members of a given online community.
Authors - Kornprom Pikulkaew, Apinantn Sumthumpruek Abstract - Investigating the emotional lives of animals is inherently complicated although the findings are rewarding in terms of conservation, psychology, and sociopsychology. In this paper, we consider the problem of animal emotion detection with the aid of deep learning using dogs' emotions in this study. In this case, four categories of emotions, angrily, joyfully, relaxed, and sadly were classified based on the best CNNs such as ResNet-50, EfficientNet, and MobileNet. Furthermore, by employing data preprocessing, data augmentation techniques, and structural explainability techniques like the Grad-CAM, we could improve how the model made critical decisions. Results confirmed that the model per-formed satisfactorily in detecting the subjects' emotions even though joyful and relaxed states were more pronounced with high levels of accuracy compared to others with emotions like sadness and anger lauded as a notable challenge, especially in discriminating attitudes that seemed too close to one another. The application of Grad-CAM was able to elaborate on the regions of interest incorporated by the model thus enhancing the explainability of the model. This paper is focused on the emergent developments in emotion detection in animals and further recommends for advancement of three-dimensional deep learning techniques, settlement of the dataset, and the introduction of more complicated explainable AI techniques such as Local Interpretable Model Explanations (LIME).
Authors - Aitor Godoy, Ismael Rodriguez, Fernando Rubio Abstract - In this paper we present a series of algorithms to calculate the power of each political party in a parliamentary system. For this purpose, it is necessary to calculate the proportion of parliamentary majorities in which their participation is necessary. The usefulness of the proposed methods is illustrated with a real case study: the Spanish electoral system. For this system, we analyze all the elections that have taken place since the establishment of democracy in the country. For each electoral process, we compare the power that each party would have if the allocation of deputies were proportional to the number of votes, and the real power it has with the current electoral system. The results obtained contradict intuitions that the Spanish population usually has about its own electoral system.
Authors - Edura Halim, Azman Mohamed, Mohamad Syazli Fathi, Zeeshan Aziz Abstract - Roads and highways largely contributes to the growth and development of a country. This paper focuses on enhancing present road safety assessment tools and techniques in line with Safe Roads, one of the pillars of the Safe System. Road Infrastructure Safety Management (RISM) is a comprehensive approach to road safety that involves a set of procedures with systematic identification, assessment, and management of risks associated with road infrastructure. This paper presents a SWOT analysis of tools used in RISM during the design stage including Road Safety Audit (RSA), Road Safety Impact Assessment (RSIA), Road Safety Screening and Appraisal Tools (RSSAT), Star Rating for Designer (SR4D) and Safe System Assessment (SSA). Conceptual framework of Intelligent Road Safety Assessment for Designer (IRSA4D) has been developed to utilized and address the findings on the SWOT analysis of RISM tools. IRSA4D application is aimed to facilitate road designers in providing proactive safety assessment and recommendations for improvement through intelligent static and dynamic assessment. The application is deemed to be valuable in assisting road designers in spite of their level of knowledge and working experiences as well as providing aid in producing the optimum and ‘best’ design during design stage.
Authors - Balendra, Neeraj Sharma, Shiru Sharma Abstract - Brain-Computer Interface (BCI) technology stands at the forefront of interdisciplinary research, merging neuroscience, engineering, and computer science to forge direct communication channels between the human brain and external devices. BCI based devices has tremendous applications in prosthetic device development. The challenges in real-time practical BCI implementation are due to the bulky models, inherent noises, artifacts and complexity of motor imagery (MI) electroencephalogram (EEG) data with inter-subject and intra-subject variabilities. To overcome these challenges, the proposed algorithm introduces a modified EEG Morlet (MEM) wavelet having a better time bandwidth product leading to detailed feature extraction with capability of natural filter for artifacts and noises introduced by eye blinking and muscle movements. Further, the proposed approach utilizes Hilbert transform to extract temporal features of analytical signal, extract their common spatial patterns, calculates the continuous wavelet transform (CWT) coefficients, arrange these coefficients at different scale for each channel, calculates the cross-correlation for each scale and observes the evolution in cross-correlation matrices at different scale with the help of customized long-short term memory (LSTM) neural network to classify MI EEG. The customized LSTM architecture had the size of 1.93MB showing the effectiveness of methodology for MI EEG classification of embedded based devices. The best classification accuracy achieved by MEM wavelet with instantaneous magnitude temporal feature was 83.78% and the comparative analysis with earlier state of the art methods showed an improvement of 1.10% in accuracy.
Authors - Ahmed D. Alharthi Abstract - Traditional proficiency and qualification criteria in crowd-based requirements engineering (CrowdRE) often fall short when identifying the most capable individuals for critical tasks such as elicitation and analysis. While crowd profiles typically account for demographic information like gender and nationality, they frequently overlook personality traits, which can significantly influence task performance. This study addresses this limitation by examining the relationship between personality traits and key requirements engineering (RE) activities within CrowdRE environments. We propose an automated system that incorporates personality profiling into task assignment processes, enhancing the precision of matching individuals to specific RE tasks. By employing data fusion techniques and decision-making algorithms, the system improves the efficiency and effectiveness of task allocation. An empirical investigation is conducted to highlight the impact of personality traits on the success of RE tasks and the need to incorporate these factors into task assignment strategies. The findings contribute to developing intelligent, human-centred collaboration technologies that optimise workflows in crowd-sourced environments. This research underscores the importance of personality traits in improving task performance and collaboration within large-scale ICT systems, aligning with the broader objective of enhancing task allocation in the context of modern collaboration technologies.
Authors - Youssef Baklouti, Tarik Echcherqaoui, Ines Abdeljaoued-Tej Abstract - This release introduces an innovative interactive chatbot designed to engage with sensitive enterprise data. Leveraging Azure Machine Learning Promptflow and Retrieval-Augmented Generation (RAG) architecture, the chatbot facilitates secure data retrieval and generation within the enterprise environment. To assess the model’s performance, we utilized over 10 query examples, providing ground-truth and context data. Evaluation strategies included system-based metrics like the F1-Score, which yielded an average score of 0.59, and AI-evaluating- AI metrics such as Coherence, Groundedness, Similarity, Fluency, and Relevance, scoring 4.50, 4.20, 4.50, 4.10, and 4.40 respectively. While AI-evaluating-AI strategies showed decent scores, the relatively low F1- Score indicates potential for improvement through fine-tuning or selecting a more suitable vector database. Overall, this interactive solution not only enhances internal operations but also demonstrates AI’s potential in automating and streamlining complex processes.
Authors - Azam Syukur Rahmatullah, Nurul Fithriyah Awaliatul Laili, Akbar Nur Aziz Abstract - This study explores the root of social pathologies in the industrial era 4.0, which young and old adolescents in Indonesia carry out. Social pathological behavior that is carried out is quite dangerous to the existence of the nation and state if not handled early. This type of research is qualitative research that is directly studied in the field to find out what causes social pathologies in this industrial era 4.0 and how to solve them. The research approach is phenomenology because it wants to thoroughly explore three (3) informants, experts in the parenting field, taken from the Yogyakarta area of Indonesia, who were interviewed directly by the researcher. The data obtained was then analyzed in depth with phenomenological analysis. The study results show several anomalous social behaviors in the industry 4.0 era caused by parenting in families that are not positive. Hence, they continue until adolescence, adulthood, and even old age. Inconsistency in parenting and failure in parenting cause children to grow into unhealthy individuals and have sick souls so that on their way, they become people who deviate from their behavior. Therefore, healthy and positive parenting must be encouraged early, including in rural and urban areas. Some positive parenting movement programs that should be implemented early on include creating a healthy parenting movement nationally, creating village and city parenting homes nationally, and making instructors and families healthy home companions nationally.
Authors - Nabaa T. Salman, Wasan A. Wali, Mohammed Lami Abstract - Solar energy is an inexhaustible source of carbon-free energy world-wide. However solar radiation determines the amount of electrical energy and current that can be produced by solar panels. In this paper, the dynamic regulation of solar radiation using Polymer Dispersed Liquid Crystal (PDLC) was studied. It regulates solar radiation that is incident on the solar cell’s surface by changing its transparency because of the applied voltage. Thus, researchers were able to obtain a variable range of solar radiation at the same time and then control the power of the solar cell produced according to the user's request. In the outdoor experiment, the behavior of filtrated radiation in daylight performance under different sky circumstances was assessed by examining solar radiation both with and without a PDLC screen. To simulate the PDLC and forecast solar radiation in real time, the researchers utilized an Adaptive Neuro-Fuzzy Inference System (ANFIS). MATLAB program utilized 3.5 W, 25 W, and 100W solar panels. PDLC transparency ranges from 5% to 83%. The results showed that PDLC overall shading on transparent\opaque states are 73% to 37% respectively at the same point where PDLC films may regulate the solar cell's output power at a pace of (39.6%, 39.5%, 42%) of the total cell power for the simulated solar panel respectively.
Authors - Miguel Angel Ruiz-Adarmes Abstract - Knowing the weather conditions at airports is of vital importance for reasons that affect the safety, efficiency, and comfort of flights. Bad weather, such as strong winds or fog, can represent a significant danger to aircraft during takeoffs, landings, and flights. Therefore, learning to predict weather behavior, based on prior information, is important. That is why this research on the prediction of weather conditions at Jorge Ch´avez Airport in Lima is presented. To do this, a set of previous data was used to which the J48, Random Forest, SVM, Bayes Net, and Neural Network algorithms were applied to identify that the Random Forest algorithm obtained the best behavior with Accuracy = 76.8004% for the training and validation process; and Accuracy = 78.5181% for the test process. As a proof of concept, a Java application was implemented.
Authors - Strahil Sokolov, Kaloyan Varlyakov, Dimitar Radev Abstract - In this paper an approach is described for improving the quality of data generated from medical screening processes based on machine learning. The designed workflow uses data acquisition from Titmus equipment, performs data preprocessing, model training and health record evaluation. We are proposing a design of a distributed system to realize this approach in order to bridge the gap between the cloud-native technologies and their usage for patient screening in rural or remote areas. The algorithm shows promising results and is suitable for implementation on Edge-AI , IoT and cloud-based medical support systems.
Authors - Marwa Mostafa Yassin, Nahla A. Belal, Aliaa Youssif Abstract - Papilledema is a medical disorder marked by the enlargement of the optic disc. Optic disk imaging is essential for the diagnosis of papilledema, as neglecting to perform this procedure can lead to fatal outcomes. This research presents a novel approach that combines deep learning with the Harris Hawks Optimization (HHO) algorithm to increase the accuracy of diagnosing and distinguishing papilledema in optical disk images. The proposed technique presented in this study focuses on optimizing the weights of the Convolutional Neural Network (CNN) model. This optimization process improves model training by using the underlying optimization principles. The technique was evaluated using the Kaggle dataset, which was made available for this purpose. The evaluation results showed that the proposed technique achieved an accuracy of 0.997%, surpassing the performance of existing techniques such as VGG16, DenseNet121, EfficientNetB0, and EfficientNetB3. The proposed model demonstrates that state-of-the-art CNN models, when paired with the HHO algorithm, can reliably diagnose real papilledema, pseudo papilledema, and normal optic discs. This could potentially save lives for patients.
Authors - Otuu Obinna Ogbonnia, Joseph Henry Anajemba, Oko-Otu Chukwuemeka N., Deepak Sahoo Abstract - Scholars have investigated the challenges of community policing (CP) in Nigeria through socio-political, economic, and cultural lenses, with none adopting a method that can reveal these challenges comprehensively. This has led to a gap in recognizing key CP problems, thereby resulting in ineffective solutions from the government, and making government services in this context less accessible and responsive to citizens. This study employed Greenhalgh’s meta-narrative approach to unveil community policing challenges that were previously overlooked in Nigerian context. Drawing from a variety of sources such as scholarly articles (ACM digital library, Science Direct), official documents, and media coverage, this study identified lack of robust technology usage, lack of citizens’ participation, citizens’ unwillingness to share information and lack of trust, accountability and transparency as major community policing challenges in Nigeria. This study contributes to a nuanced understanding of the challenges hindering the successful implementation of CP in Nigeria, highlights the implications of these challenges on the overall security landscape, and offers directions to policymakers and relevant government agencies, providing insights to the design of technological solutions for community policing in Nigeria.
