E-Learning is an emerging dominant phenomenon in education, making the development of robust models that can accurately represent the dynamic behavior of learners in MOOCs even more critical. In this article, we propose the Temporal Sparse Attention-Gated Recurrent Unit (TSA-GRU), a novel deep learning framework that combines TSA with a sequential encoder based on the GRU. This hybrid model effectively reconstructs student response times and learning trajectories with high fidelity by leveraging tthe emporal embeddings of instructional and feedback activities. By dynamically filtering noise from student interactions, TSA-GRU generates context-aware representations that seamlessly integrate both short-term fluctuations and long-term learning patterns. Empirical evaluation on the 2009–2010 ASSISTments dataset demonstrates that TSA-GRU achieved a test accuracy of 95.60% and a test loss of 0.0209, outperforming Modular Sparse Attention-Gated Recurrent Unit (MSA-GRU), Bayesian Knowledge Tracing (BKT), Performance Factors Analysis (PFA), and TSA in the same experimental design. TSA-GRU converged in five training epochs; thus, while TSA-GRU is demonstrated to have strong predictive performance for knowledge tracing tasks, these findings are specific to the conducted dataset and should not be implicitly regarded as conclusive for all data. More statistical validation through five-fold cross-validation, confidence intervals, and paired t-tests have confirmed the robustness, consistency, and statistically significant superiority of TSA-GRU over the baseline model MSA-GRU. TSA-GRU’s scalability and capacity to incorporate a temporal dimension of knowledge can make it acceptably well-positioned to analyze complex learner behaviors and plan interventions for adaptive learning in computerized learning systems.

The field of autonomous robotics is progressing rapidly, with research moving toward developing systems capable of moving without direct human control and learning without human intervention. One of the problems requiring an efficient and sustainable solution is ensuring the smooth and safe navigation of robots between obstacles. In this study, a new path planning approach is developed, integrating particle swarm optimization (PSO) and artificial potential field (APF) algorithms to assist the mobile robot in navigating an area with static and dynamic obstacles. The robot moves independently while routing dynamically and avoiding obstacles. To evaluate its adaptive ability to a changing environment, we continuously calculate the shortest distance between two points and dynamically adjust the path to avoid obstacles during replanning, path recalculation, and robot position adjustment to ensure efficient and safe navigation. Different scenarios are tested to evaluate our approach, including different environmental conditions and obstacle configurations. Experimental results show that our method reduces the path length by 18%, the obstacle avoidance efficiency by 90%, and the success rate by 85% in dynamic environments. In addition, PSO-APF reduces computation time, demonstrating better capacity and efficiency.

Artificial Intelligence (AI) and computer-aided diagnosis (CAD) have revolutionised various aspects of modern life, particularly in the medical domain. These technologies enable efficient solutions for complex challenges, such as accurately segmenting brain tumour regions, which significantly aid medical professionals in monitoring and treating patients. This research focuses on segmenting glioma brain tumour lesions in MRI images by analysing them at the pixel level. The aim is to develop a deep learning-based approach that enables ensemble learning to achieve precise and consistent segmentation of brain tumours. While many studies have explored ensemble learning techniques in this area, most rely on aggregation functions like the Weighted Arithmetic Mean (WAM) without accounting for the interdependencies between classifier subsets. To address this limitation, the Choquet integral is employed for ensemble learning, along with a novel evaluation framework for fuzzy measures. This framework integrates coalition game theory, information theory, and Lambda fuzzy approximation. Three distinct fuzzy measure sets are computed using different weighting strategies informed by these theories. Based on these measures, three Choquet integrals are calculated for segmenting different components of brain lesions, and their outputs are subsequently combined. The BraTS-2020 online validation dataset is used to validate the proposed approach. Results demonstrate superior performance compared with several recent methods, achieving Dice Similarity Coefficients of 0.896, 0.851, and 0.792 and 95% Hausdorff distances of 5.96 mm, 6.65 mm, and 20.74 mm for the whole tumour, tumour core, and enhancing tumour core, respectively.

The progress of the Big Data technology in the different domains continues to reveal, day after day, so many issues affecting this field. The data presentation is a part of these issues, because it is directly related to the user that requires visualization to facilitate the data collection as well as the analysis and the utilization of results as needed. However, implementing visualization requires a low latency. Therefore, many problems regarding the Big Data major characteristics, namely volume, variety and velocity arise and should be addressed. Knowing that each Big Data domain has specific peculiarities, it remains far from thinking of a generic solution to the visualization problem. This latter is manifested in different kinds. One of these kinds that is prevalent in Big Data domains is hierarchical visualization, which is based on presenting data by levels, while respecting the relations between them. In this paper, we propose an architecture composed of modules that cover the Big Data visualization process, taking into account the hierarchical relations between data. In order to valorise our proposal, we implement a prototype that supports the modules composing our architecture which are developed using different tools such as Scrapy and D3js libraries. The experimentation, which was carried out in the educational domain, showed its applicability to real systems.

E-Learning is an emerging dominant phenomenon in education, making the development of robust models that can accurately represent the dynamic behavior of learners in MOOCs even more critical. In this article, we propose Temporal Sparse Attention- Gated Recurrent Unit (TSA-GRU), which is a new framework in deep learning that combines TSA with a sequential encoder based on a GRU. This hybrid model effectively reconstructs student response times and learning trajectories with high fidelity by leveraging temporal embedding of instructional and feedback activities. By dynamically filtering noise from student interactions, TSA-GRU generates context-aware representations that seamlessly integrate both short-term fluctuations and long-term learning patterns. Empirical tests based on a large-scale randomized MOOC dataset demonstrate that TSA-GRU has superior accuracy and performance than Modular Sparse Attention with Gated Recurrent Units (MSA-GRU), Bayesian Knowledge Tracing (BKT), Performance Factors Analysis (PFA), and the Temporal Sparse Attention (TSA) model alone. TSA-GRU achieves an impressive 95.60% accuracy on the test with loss of 0.0209, which is also demonstrated by its rapid convergence within five training epochs, implying strength and efficiency compared to others. These results offer evidence that TSA-GRU has the potential to contribute as a state-of-the-art model for adaptive learning systems that incorporate timely and personalized interventions, and that its scalability and efficiency captures the complexity of learner behavior in order to solve issues associated to modeling learner behavior in digital education settings and predictive efficacy in modern e-learning contexts.