Big Data and Cognitive Computing

Text summarization is a complex and essential task in natural language processing (NLP) research, focused on extracting the most important information from a document. This study focuses on the Extractive and Abstractive approaches of Bengali Text Summarization (BTS). With the breakthrough advancements in deep learning, summarization is no longer a major challenge for English, given the availability of extensive resources dedicated to this global language. However, the Bengali language remains underexplored. Hence, in this work, a comprehensive review has been conducted on BTS research from 2007 to 2023, analyzing trends, datasets, preprocessing techniques, methodologies, evaluations, and challenges. Leveraging 106 journal and conference papers, this review offers insights into emerging topics and trends in Bengali Abstractive summarization. The review has been augmented with experiments using transformer models from Hugging Face and publicly available datasets to assess the Rouge score accuracy for Abstractive summarization. The extensive literature review conducted in this study reveals that before the advent of transformers, LSTM (Long Short-Term Memory) models were the dominant deep learning approach for text summarization across various languages. For transformers, one of the key datasets utilized was XL-SUM with the MT5 model emerging as the best performer among various contemporary multilingual models. These findings contribute to understanding the contemporary techniques and challenges in BTS. Furthermore, recommendations are made to guide future research endeavors, aiming to provide valuable insights and directions for researchers in this field. Full article

The decision-making process to rule R&D relies on information related to current trends in particular research areas. In this work, we investigated how one can use large language models (LLMs) to transfer the dataset and its annotation from one language to another. This is crucial since sharing knowledge between different languages could boost certain underresourced directions in the target language, saving lots of effort in data annotation or quick prototyping. We experiment with English and Russian pairs, translating the DEFT (Definition Extraction from Texts) corpus. This corpus contains three layers of annotation dedicated to term-definition pair mining, which is a rare annotation type for Russian. The presence of such a dataset is beneficial for the natural language processing methods of trend analysis in science since the terms and definitions are the basic blocks of any scientific field. We provide a pipeline for the annotation transfer using LLMs. In the end, we train the BERT-based models on the translated dataset to establish a baseline. Full article

Deep learning applications for Edge Intelligence (EI) face challenges in achieving high model performance while maintaining computational efficiency, particularly under varying image orientations and perspectives. This study investigates the synergy of multi-backbone (MB) configurations and Synchronized Multi Augmentation (SMA) to address these challenges by leveraging diverse input representations and spatial transformations. SMA employs synchronously augmented input data across MBs during training, thereby improving feature extraction across diverse representations. The outputs provided by these MBs are merged through different fusion strategies: Averaging Fusion with aggregation of predictions and Dense Fusion with integration of features via a fully connected neural network. It aims to increase model accuracy on previously unseen input data and to reduce computational requirements by minimizing neural network size, particularly advantageous for EI systems characterized by the limited computing resources. This study employed MBs with the MobileNetV3 architecture and the CIFAR-10 dataset to investigate the impact of SMA techniques and different fusion strategies on model robustness and performance. SMA techniques were applied to simulate diverse image orientations, and MB architectures were tested with Averaging and Dense fusion strategies to assess their ability to learn diverse feature representations and improve robustness. The experiments revealed that models augmented with SMA outperformed the baseline MobileNetV3 on modified datasets, achieving higher robustness to orientation variations. Models with Averaging fusion exhibited the most stable performance across datasets, while Dense fusion achieved the highest metrics under specific conditions. Results indicate that SMAs incorporating image transformation adjustments, such as rotation, significantly enhance generalization across varying orientation conditions. This approach enables the production of more stable results using the same pretrained weights in real-world applications by configuring Image Signal Processing (ISP) to effectively use SMA. The findings encourage further exploration of SMA techniques in conjunction with diverse camera sensor configurations and ISP settings to optimize real-world deployments. Full article

As AI becomes indispensable in healthcare, its vulnerability to adversarial attacks demands serious attention. Even minimal changes to the input data can mislead Deep Learning (DL) models, leading to critical errors in diagnosis and endangering patient safety. In this study, we developed an optimized Multi-layer Perceptron (MLP) model for breast cancer classification and exposed its cybersecurity vulnerabilities through a real-world-inspired adversarial attack. Unlike prior studies, we conducted a quantitative evaluation on the impact of a Fast Gradient Sign Method (FGSM) attack on an optimized DL model designed for breast cancer detection to demonstrate how minor perturbations reduced the model’s accuracy from 98% to 53%, and led to a substantial increase in the classification errors, as revealed by the confusion matrix. Our findings demonstrate how an adversarial attack can significantly compromise the performance of a healthcare AI model, underscoring the importance of aligning AI development with cybersecurity readiness. This research highlights the demand for designing resilient AI by integrating rigorous cybersecurity practices at every stage of the AI development lifecycle, i.e., before, during, and after the model engineering to prioritize the effectiveness, accuracy, and safety of AI in real-world healthcare environments. Full article

In this article, for the first time on this topic, we analyze the historical color palettes of Renaissance oil paintings by using machine-learning methods and digital images. Our work has two main parts: we collect data on their historical color palettes and then use machine learning to predict the original colors of paintings. This model studies color ratios, enhancement levels, symbolic meanings, and historical records. It looks at key colors, measures their relationships, and learns how they have changed. The main contributions of this work are as follows: (i) we develop a model that predicts a painting’s original color palette based on multiple factors, such as the color ratios and symbolic meanings, and (ii) we propose a framework for using cognitive computing tools to recover the original colors of historical artworks. This helps us to rediscover lost emotional and cultural details. Full article

Effective student feedback is fundamental to enhancing learning outcomes in higher education. While traditional assessment methods emphasise both achievements and development areas, the process remains time-intensive for educators. This research explores the application of cognitive computing, specifically open-source Large Language Models (LLMs) Mistral-7B and CodeLlama-7B, to streamline feedback generation for student reports containing both Python programming elements and English narrative content. The findings indicate that these models can provide contextually appropriate feedback on both technical Python coding and English specification and documentation. They effectively identified coding weaknesses and provided constructive suggestions for improvement, as well as insightful feedback on English language quality, structure, and clarity in report writing. These results contribute to the growing body of knowledge on automated assessment feedback in higher education, offering practical insights for institutions considering the implementation of open-source LLMs in their workflows. There are around 22 thousand assessment submissions per year in the School of Computer Science, which is one of eight schools in the Faculty of Engineering and Physical Sciences, which is one of seven faculties in the University of Leeds, which is one of one hundred and sixty-six universities in the UK, so there is clear potential for our methods to scale up to millions of assessment submissions. This study also examines the limitations of current approaches and proposes potential enhancements. The findings support a hybrid system where cognitive computing manages routine tasks and educators focus on complex, personalised evaluations, enhancing feedback quality, consistency, and efficiency in educational settings. Full article

The intersection of medical image classification and deep learning has garnered increasing research interest, particularly in the context of breast tumor detection using ultrasound images. Prior studies have predominantly focused on image classification, segmentation, and feature extraction, often assuming that the input images, whether sourced from healthcare professionals or individuals, are valid and relevant for analysis. To address this, we propose an initial binary classification filter to distinguish between relevant and irrelevant images, ensuring only meaningful data proceeds to subsequent analysis. However, the primary focus of this study lies in investigating the performance of a hierarchical two-tier classification architecture compared to a traditional flat three-class classification model, by employing a well-established breast ultrasound images dataset. Specifically, we explore whether sequentially breaking down the problem into binary classifications, first identifying normal versus tumorous tissue and then distinguishing benign from malignant tumors, yields better accuracy and robustness than directly classifying all three categories in a single step. Using a range of evaluation metrics, the hierarchical architecture demonstrates notable advantages in certain critical aspects of model performance. The findings of this study provide valuable guidance for selecting the optimal architecture for the final model, facilitating its seamless integration into a web application for deployment. These insights are further anticipated to advance future algorithm development and broaden the potential of the research applicability across diverse fields. Full article

A trajectory is a set of time-stamped locations of a moving object usually recorded by GPS sensors. Today, an abundance of these data is available. These large quantities of data need to be analyzed to determine patterns and associations of interest to business analysts. In this paper, a formal model of trajectories is proposed, which focuses on the aggregation of semantic attributes. These attributes can be associated by the analyst to different structural elements of a trajectory (either to the points, to the edges, or to the entire trajectory). The model allows the analyst to specify not only these semantic attributes, but also to specify for each semantic attribute the set of aggregation operators (SUM, AVG, MAX, MIN, etc.) that the analyst considers appropriate to be applied to the attribute in question. The concept of PAV (package of aggregate values) is also introduced and formalized. PAVs can help identify patterns in traffic, tourism, migrations, among other fields. Experiments with real data about trajectories of people revealed interesting findings about the way people move and showed the expediency and usefulness of the proposal. The contributions in this work provide a foundation for future research in developing trajectory applications including analysis and visualization of trajectory aggregated data based on formal grounds. Full article

Deepfake technology poses significant threats in various domains, including politics, cybersecurity, and social media. This study uses the golden frame selection technique to present a neural network ensemble method for deepfake classification. The proposed approach optimizes computational resources by extracting the most informative video frames, improving detection accuracy. We integrate multiple deep learning models, including ResNet50, EfficientNetB0, Xception, InceptionV3, and Facenet, with an XGBoost meta-model for enhanced classification performance. Experimental results demonstrate a 91% accuracy rate, outperforming traditional deepfake detection models. Additionally, feature importance analysis using Grad-CAM highlights how different architectures focus on distinct facial regions, enhancing overall model interpretability. The findings contribute to of robust and efficient deepfake detection techniques, with potential applications in digital forensics, media verification, and cybersecurity. Full article

Ensuring the trustworthiness of data used in real-time analytics remains a critical challenge in smart city monitoring and decision-making. This is because the traditional data validation methods are insufficient for handling the dynamic and heterogeneous nature of Internet of Things (IoT) data streams. This paper describes a semantic IoT streaming data validation approach to provide a semantic IoT data model and process IoT streaming data with the semantic stream processing systems to check the quality requirements of IoT streams. The proposed approach enhances the understanding of smart city data while supporting real-time, data-driven decision-making and monitoring processes. A publicly available sensor dataset collected from a busy road in Milan city is constructed, annotated and semantically processed by the proposed approach and its architecture. The architecture, built on a robust semantic-based system, incorporates a reasoning technique based on forward rules, which is integrated within the semantic stream query processing system. It employs serialized Resource Description Framework (RDF) data formats to enhance stream expressiveness and enables the real-time validation of missing and inconsistent data streams within continuous sliding-window operations. The effectiveness of the approach is validated by deploying multiple RDF stream instances to the architecture before evaluating its accuracy and performance (in terms of reasoning time). The approach underscores the capability of semantic technology in sustaining the validation of IoT streaming data by accurately identifying up to 99% of inconsistent and incomplete streams in each streaming window. Also, it can maintain the performance of the semantic reasoning process in near real time. The approach provides an enhancement to data quality and credibility, capable of providing near-real-time decision support mechanisms for critical smart city applications, and facilitates accurate situational awareness across both the application and operational levels of the smart city. Full article

Recent advances in computer science and geomatics have enabled the digitalization of complex two-dimensional and three-dimensional spatial environments and the sharing of geospatial data on the web. Simultaneously, the widespread adoption of Internet of Things (IoT) technology has facilitated the rapid deployment of low-cost sensor networks in various scientific applications. The integration of real-time IoT data acquisition in 3D urban environments lays the foundation for the development of Urban Digital Twins. This work proposes a possible low-cost solution as a sample of a structure for 3D digital twinning on the web, presenting a case study related to weather monitoring analysis. Specifically, an indoor-outdoor environmental conditions monitoring system integrated with 3D geospatial data on a 3D WebGIS platform was developed. This solution can be considered as a first step for monitoring human and environmental wellness within a geospatial analysis system that integrates several open-source modules that provide different kinds of information (geospatial data, 3D models, and IoT acquisition). The structure of this system can be valuable for municipalities and private stakeholders seeking to conduct environmental geospatial analysis using cost-effective solutions. Full article

Recommender systems aspire to provide users with recommendations that have a high probability of being accepted. This is accomplished by producing rating predictions for products that the users have not evaluated, and, afterwards, the products with the highest prediction scores are recommended to them. Collaborative filtering is a popular recommender system technique which generates rating prediction scores by blending the ratings that users with similar preferences have previously given to these products. However, predictions may entail errors, which will either lead to recommending products that the users would not accept or failing to recommend products that the users would actually accept. The first case is considered much more critical, since the recommender system will lose a significant amount of reliability and consequently interest. In this paper, after performing a study on rating prediction confidence factors in collaborative filtering, (a) we introduce the concept of prediction reliability classes, (b) we rank these classes in relation to the utility of the rating predictions belonging to each class, and (c) we present a collaborative filtering recommendation algorithm which exploits these reliability classes for prediction formulation. The efficacy of the presented algorithm is evaluated through an extensive multi-parameter evaluation process, which demonstrates that it significantly enhances recommendation quality. Full article

The Maltese Islands, situated at the centre of the Mediterranean basin, are recognised as a climate change hotspot. This study utilises projected changes in temperature and precipitation derived from the World Climate Research Program (WCRP) and analyses outputs from six coupled model intercomparison project phase 5 (CMIP5) models under two Representative Concentration pathways (RCPs). Through statistical and spatial analysis, the study demonstrates that climate change will have significant adverse effects on Maltese agriculture. Regardless of the RCP scenario considered, projections indicate a substantial increase in temperature and a decline in precipitation, exacerbating aridity and intensifying heat stress. These changes are expected to reduce soil moisture availability and challenge traditional agricultural practices. The study identifies the Western District as a relatively more favourable area for crop cultivation due to its comparatively lower temperatures, whereas the Northern and South Eastern peripheries are projected to experience more severe heat stress. Adaptation strategies, including the selection of heat-tolerant crop varieties such as Tetyda and Finezja, optimised water management techniques, and intercropping practices, are proposed to enhance agricultural resilience. This study is among the few comprehensive assessments of bioclimatic and physical factors affecting Maltese agriculture and highlights the urgent need for targeted adaptation measures to safeguard food production in the region. Full article

The development of cybernetic organisms—autonomous systems capable of self-regulation and dynamic environmental interaction—requires innovations in both energy efficiency and computational adaptability. This study explores the integration of bio-inspired liquid flow batteries and neuromorphic computing architectures to enable real-time learning and power optimization in autonomous robotic systems. Liquid-based energy storage systems, modeled after vascular networks, offer distributed energy management, reducing power bottlenecks and improving resilience in long-duration operations. Complementing this, neuromorphic computing architectures, including memristor-based processors and spiking neural networks (SNNs), enhance computational efficiency while minimizing energy consumption. By integrating these adaptive energy and computing systems, robots can dynamically allocate power and processing resources based on real-time demands, bridging the gap between biological and artificial intelligence. This study evaluates the feasibility of integrating these technologies into robotic platforms, assessing power demands, storage capacity, and operational scalability. While flow batteries and neuromorphic computing show promise in reducing latency and energy constraints, challenges remain in electrolyte stability, computational framework standardization, and real-world implementation. Future research must focus on hybrid computing architectures, self-regulating energy distribution, and material optimizations to enhance the adaptability of cybernetic organisms. By addressing these challenges, this study outlines a roadmap for reimagining robotics through cybernetic principles, paving the way for applications in healthcare, industrial automation, space exploration, and adaptive autonomous systems in dynamic environments. Full article

Word games are of great importance in the acquisition of vocabulary and letter recognition among children, usually between the ages of 3 and 13, boosting their memory, word retention, spelling, and cognition. Despite the importance of these games, little attention has been paid to the development of word games for low-resource or highly morphologically constructed languages. This study develops an Advanced Cubic-oriented Game (ACG) model by using a character-level N-gram technique and statistics, commonly known as the matching letter game, wherein a player forms words using a given number of cubes with letters on each of its sides. The main objective of this study is to find out the optimal number of letter cubes while maintaining the overall coverage. Comprehensive experiments on 12 datasets (from low-resource and high-resource languages) incorporating morphological features were conducted to form 3–5-letter words using 7–8 cubes and a special case of forming 6–7-letter words using 8–9 cubes. Experimental evaluations show that the ACG model achieved reasonably high results in terms of average total coverage, with 89.5% for 3–5-letter words using eight cubes and 79.7% for 6–7-letter words using nine cubes over 12 datasets. The ACG model obtained over 90% coverage for Uzbek, Turkish, English, Slovenian, Spanish, French, and Malaysian when constructing 3–5-letter words using eight cubes. Full article

Location-based social networks (LBSNs) leverage geo-location technologies to connect users with places, events, and other users nearby. Using GPS data, platforms like Foursquare enable users to check into locations, share their locations, and receive location-based recommendations. A significant research gap in LBSNs lies in the limited exploration of users’ tendencies to withhold certain location data. While existing studies primarily focus on the locations users choose to disclose and the activities they attend, there is a lack of research on the hidden or intentionally omitted locations. Understanding these concealed patterns and integrating them into predictive models could enhance the accuracy and depth of location prediction, offering a more comprehensive view of user mobility behavior. This paper solves this gap by proposing an Associative Hidden Location Trajectory Prediction model (AHLTP) that leverages user trajectories to infer unchecked locations. The FP-growth mining technique is used in AHLTP to extract frequent patterns of check-in locations, combined with machine-learning methods such as K-nearest-neighbor, gradient-boosted-trees, and deep learning to classify hidden locations. Moreover, AHLTP uses association rule mining to derive the frequency of successive check-in pairs for the purpose of hidden location prediction. The proposed AHLTP integrated with the machine-learning models classifies the data effectively, with the KNN attaining the highest accuracy at 98%, followed by gradient-boosted trees at 96% and deep learning at 92%. Comparative study using a real-world dataset demonstrates the model’s superior accuracy compared to state-of-the-art approaches. Full article

This study investigated the characteristics and functionalities of China’s High-Speed Railway (HSR) network based on Complex Network Theory (CNT) and Graph Convolutional Networks (GCN). First, complex network analysis was applied to provide insights into the network’s fundamental characteristics, such as small-world properties, efficiency, and robustness. Then, this research developed three novel GCN models to identify key nodes, detect community structures, and predict new links. Findings from the complex network analysis revealed that China’s HSR network exhibits a typical small-world property, with a degree distribution that follows a log-normal pattern rather than a power law. The global efficiency indicator suggested that stations are typically connected through direct routes, while the local efficiency indicator showed that the network performs effectively within local areas. The robustness study indicated that the network can quickly lose connectivity if key nodes fail, though it showed an ability initially to self-regulate and has partially restored its structure after disruption. The GCN model for key node identification revealed that the key nodes in the network were predominantly located in economically significant and densely populated cities, positively contributing to the network’s overall efficiency and robustness. The community structures identified by the integrated GCN model highlight the economic and social connections between official urban clusters and the communities. Results from the link prediction model suggest the necessity of improving the long-distance connectivity across regions. Future work will explore the network’s socio-economic dynamics and refine and generalise the GCN models. Full article

The emergence of Multimodal Large Language Models (MLLMs) has made methods of artificial intelligence accessible to the general public in a conversational way. It offers tools for the automated visual assessment of the quality of a built environment for professionals of urban planning without requiring specific technical knowledge on computing. We investigated the capability of MLLMs to perceive urban environments based on images and textual prompts. We compared the outputs of several popular models—ChatGPT, Gemini and Grok—to the visual assessment of experts in Architecture, Engineering and Construction (AEC) in the context of a real estate construction project. Our analysis was based on subjective attributes proposed to characterize various aspects of a built environment. Four urban identities served as case studies, set in a virtual environment designed using professional 3D models. We found that there can be an alignment between human and AI evaluation on some aspects such as space and scale and architectural style, and more general accordance in environments with vegetation. However, there were noticeable differences in response patterns between the AIs and AEC experts, particularly concerning subjective aspects such as the general emotional resonance of specific urban identities. It raises questions regarding the hallucinations of generative AI where the AI invents information and behaves creatively but its outputs are not accurate. Full article

College enrollment has long been recognized as a critical pathway to better employment prospects and improved economic outcomes. However, the overall enrollment rates have declined in recent years, and students with a lower socioeconomic status (SES) or those from disadvantaged backgrounds remain significantly underrepresented in higher education. To investigate the factors influencing college enrollment among low-SES high school students, this study analyzed data from the High School Longitudinal Study of 2009 (HSLS:09) using five widely used machine learning algorithms. The sample included 5223 ninth-grade students from lower socioeconomic backgrounds (51% female; M_age = 14.59) whose biological parents or stepparents completed a parental questionnaire. The results showed that, among all five classifiers, the random forest algorithm achieved the highest classification accuracy at 67.73%. Additionally, the top three predictors of enrollment in 2-year or 4-year colleges were students’ overall high school GPA, parental educational expectations, and the number of close friends planning to attend a 4-year college. Conversely, the most important predictors of non-enrollment were high school GPA, parental educational expectations, and the number of close friends who had dropped out of high school. These findings advance our understanding of the factors shaping college enrollment for low-SES students and highlight two important factors for intervention: improving students’ academic performance and fostering future-oriented goals among their peers and parents. Full article