Search Results (3,822)

Recent advancements in large language models (LLMs) have significantly enhanced the effectiveness of Retrieval-Augmented Generation (RAG) systems. This study focuses on the development and evaluation of a domain-specific AI chatbot designed to support international student admissions by leveraging LLM-based RAG pipelines. We implement and compare multiple pipeline configurations, combining retrieval methods (e.g., Dense, MMR, Hybrid), chunking strategies (e.g., Semantic, Recursive), and both open-source and commercial LLMs. Dual evaluation datasets of LLM-generated and human-tagged QA sets are used to measure answer relevancy, faithfulness, context precision, and recall, alongside heuristic NLP metrics. Furthermore, latency analysis across different RAG stages is conducted to assess deployment feasibility in real-world educational environments. Results show that well-optimized open-source RAG pipelines can offer comparable performance to GPT-4o while maintaining scalability and cost-efficiency. These findings suggest that the proposed chatbot system can provide a practical and technically sound solution for international student services in resource-constrained academic institutions. Full article

In distance learning environments, learner engagement directly impacts attention, motivation, and academic performance. Signs of fatigue, negative affect, or critical remarks can warn of growing disengagement and potential dropout. However, most existing approaches rely on a single modality, visual or text-based, without providing a general view of learners’ cognitive and affective states. We propose a multimodal system that integrates three complementary analyzes: (1) a CNN-LSTM model augmented with warning signs such as PERCLOS and yawning frequency for fatigue detection, (2) facial emotion recognition by EmoNet and an LSTM to handle temporal dynamics, and (3) sentiment analysis of feedback by a fine-tuned BERT model. It was evaluated on three public benchmarks: DAiSEE for fatigue, AffectNet for emotion, and MOOC Review (Coursera) for sentiment analysis. The results show a precision of 88.5% for fatigue detection, 70% for emotion detection, and 91.5% for sentiment analysis. Aggregating these cues enables an accurate identification of disengagement periods and triggers individualized pedagogical interventions. These results, although based on independently sourced datasets, demonstrate the feasibility of an integrated approach to detecting disengagement and open the door to emotionally intelligent learning systems with potential for future work in real-time content personalization and adaptive learning assistance. Full article

Lighting Systems (LSs) are a key component of modern cities. Across the years, thousands of articles have been published on this topic; nevertheless, a map of the state of the art of the extant literature is lacking. The present review reports on an analysis of the network of the co-occurrences of the authors’ keywords from 12,148 Scopus-indexed articles on LSs published between 1995 and 2024. This review addresses the following research questions: (RQ1) What are the major topics explored by scholars in connection with LSs within the 1995–2024 time frame? (RQ2) How do they group together? The investigation leveraged VOSviewer, an open-source software largely used for performing bibliometric analyses. The number of thematic clusters returned by VOSviewer was determined by the value of the minimum number of occurrences needed for the authors’ keywords to be admitted into the business analysis. If such a number is not properly chosen, the consequence is a set of clusters that do not represent meaningful patterns of the input dataset. In the present study, to overcome this issue, the threshold value balanced the score of four independent clustering validity indices against the authors’ judgment of a meaningful partition of the input dataset. In addition, our review delved into the impact that the use/non-use of a thesaurus of the authors’ keywords had on the number and composition of the thematic clusters returned by VOSviewer and, ultimately, on how this choice affected the correctness of the interpretation of the clusters. The study adhered to a well-known protocol, whose implementation is reported in detail. Thus, the workflow is transparent and replicable. Full article

Landslide displacement prediction is crucial for disaster mitigation, yet traditional methods often fail to capture the complex, non-stationary spatiotemporal dynamics of slope evolution. This study introduces an enhanced prediction framework that integrates multi-scale signal processing with dynamic, geology-aware graph modeling. The proposed methodology first employs the Maximum Overlap Discrete Wavelet Transform (MODWT) to denoise raw Global Navigation Satellite System (GNSS)-monitored displacement time series data, enhancing the underlying deformation features. Subsequently, a geology-aware graph is constructed, using the temporal correlation of displacement series as a practical proxy for physical relatedness between monitoring nodes. The framework’s core innovation lies in a dynamic graph optimization model with low-rank constraints, which adaptively refines the graph topology to reflect time-varying inter-sensor dependencies driven by factors like mining activities. Experiments conducted on a real-world dataset from an active open-pit mine demonstrate the framework’s superior performance. The DCRNN-proposed model achieved the highest accuracy among eight competing models, recording a Root Mean Square Error (RMSE) of 2.773 mm in the Vertical direction, a 39.1% reduction compared to its baseline. This study validates that the proposed dynamic graph optimization approach provides a robust and significantly more accurate solution for landslide prediction in complex, real-world engineering environments. Full article

Rural regions face significant barriers to adopting IoT technologies, due to limited connectivity, energy constraints, and poor technical infrastructure. While urban environments benefit from advanced digital systems and cloud services, rural areas often lack the necessary conditions to deploy and evaluate secure and autonomous IoT solutions. To help overcome this gap, this paper presents the Smart Rural IoT Lab, a modular and reproducible testbed designed to replicate the deployment conditions in rural areas using open-source tools and affordable hardware. The laboratory integrates long-range and short-range communication technologies in six experimental scenarios, implementing protocols such as MQTT, HTTP, UDP, and CoAP. These scenarios simulate realistic rural use cases, including environmental monitoring, livestock tracking, infrastructure access control, and heritage site protection. Local data processing is achieved through containerized services like Node-RED, InfluxDB, MongoDB, and Grafana, ensuring complete autonomy, without dependence on cloud services. A key contribution of the laboratory is the generation of structured datasets from real network traffic captured with Tcpdump and preprocessed using Zeek. Unlike simulated datasets, the collected data reflect communication patterns generated from real devices. Although the current dataset only includes benign traffic, the platform is prepared for future incorporation of adversarial scenarios (spoofing, DoS) to support AI-based cybersecurity research. While experiments were conducted in an indoor controlled environment, the testbed architecture is portable and suitable for future outdoor deployment. The Smart Rural IoT Lab addresses a critical gap in current research infrastructure, providing a realistic and flexible foundation for developing secure, cloud-independent IoT solutions, contributing to the digital transformation of rural regions. Full article

Flocculants are indispensable in water and wastewater treatment, enabling the aggregation and removal of suspended particles, colloids, and emulsions. However, the conventional development and application of flocculants rely heavily on empirical methods, which are time-consuming, resource-intensive, and environmentally problematic due to issues such as sludge production and chemical residues. Recent advances in machine learning (ML) have opened transformative avenues for the design, optimization, and intelligent application of flocculants. This review systematically examines the integration of ML into flocculant research, covering algorithmic approaches, data-driven structure–property modeling, high-throughput formulation screening, and smart process control. ML models—including random forests, neural networks, and Gaussian processes—have successfully predicted flocculation performance, guided synthesis optimization, and enabled real-time dosing control. Applications extend to both synthetic and bioflocculants, with ML facilitating strain engineering, fermentation yield prediction, and polymer degradability assessments. Furthermore, the convergence of ML with IoT, digital twins, and life cycle assessment tools has accelerated the transition toward sustainable, adaptive, and low-impact treatment technologies. Despite its potential, challenges remain in data standardization, model interpretability, and real-world implementation. This review concludes by outlining strategic pathways for future research, including the development of open datasets, hybrid physics–ML frameworks, and interdisciplinary collaborations. By leveraging ML, the next generation of flocculant systems can be more effective, environmentally benign, and intelligently controlled, contributing to global water sustainability goals. Full article

The growth and sustainability of today’s global economy heavily relies on smooth maritime operations. The increasing security concerns to marine environments pose complex security challenges, such as smuggling, illegal fishing, human trafficking, and environmental threats, for traditional surveillance methods due to their limitations. Artificial intelligence (AI), particularly deep learning, has offered strong capabilities for automating object detection, anomaly identification, and situational awareness in maritime environments. In this paper, we have reviewed the state-of-the-art deep learning models mainly proposed in recent literature (2020–2025), including convolutional neural networks, recurrent neural networks, Transformers, and multimodal fusion architectures. We have highlighted their success in processing diverse data sources such as satellite imagery, AIS, SAR, radar, and sensor inputs from UxVs. Additionally, multimodal data fusion techniques enhance robustness by integrating complementary data, yielding more detection accuracy. There still exist challenges in detecting small or occluded objects, handling cluttered scenes, and interpreting unusual vessel behaviours, especially under adverse sea conditions. Additionally, explainability and real-time deployment of AI models in operational settings are open research areas. Overall, the review of existing maritime literature suggests that deep learning is rapidly transforming maritime domain awareness and response, with significant potential to improve global maritime security and operational efficiency. We have also provided key datasets for deep learning models in the maritime security domain. Full article

According to the latest Verizon DBIR report, credential abuse, including password reuse and human factors in password creation, remains the leading attack vector. It was revealed that most users change their passwords only when they forget them, and 35% of respondents find mandatory password rotation policies inconvenient. These findings highlight the importance of combining technical solutions with user-focused education to strengthen password security. In this research, the “human factor in the creation of usernames and passwords” is considered a vulnerability, as identifying the patterns or rules used by users in password generation can significantly reduce the number of possible combinations that attackers need to try in order to gain access to personal data. The proposed method based on an LSTM model operates at a character level, detecting recurrent structures and generating generalized masks that reflect the most common components in password creation. Open datasets of 31,000 compromised passwords from real-world leaks were used to train the model and it achieved over 90% test accuracy without signs of overfitting. A new method of evaluating the individual password creation habits of users and automatically fetching context-rich keywords from a user’s public web and social media footprint via a keyword-extraction algorithm is developed, and this approach is incorporated into a web application that allows clients to locally fine-tune an LSTM model locally, run it through ONNX, and carry out all inference on-device, ensuring complete data confidentiality and adherence to privacy regulations. Full article

High-quality, openly accessible clinical datasets remain a significant bottleneck in advancing both research and clinical applications within medical artificial intelligence. Case reports, often rich in multimodal clinical data, represent an underutilized resource for developing medical AI applications. We present an enhanced version of MultiCaRe, a dataset derived from open-access case reports on PubMed Central. This new version addresses the limitations identified in the previous release and incorporates newly added clinical cases and images (totaling 93,816 and 130,791, respectively), along with a refined hierarchical taxonomy featuring over 140 categories. Image labels have been meticulously curated using a combination of manual and machine learning-based label generation and validation, ensuring a higher quality for image classification tasks and the fine-tuning of multimodal models. To facilitate its use, we also provide a Python package for dataset manipulation, pretrained models for medical image classification, and two dedicated websites. The updated MultiCaRe dataset expands the resources available for multimodal AI research in medicine. Its scale, quality, and accessibility make it a valuable tool for developing medical AI systems, as well as for educational purposes in clinical and computational fields. Full article

News recommendation systems must simultaneously address multiple challenges, including dynamic user interest modeling, nonlinear popularity patterns, and diversity recommendation in cold-start scenarios. We present a Popularity-Aware Dynamic Multi-Perspective Fusion Network (PAD-MPFN) that innovatively integrates three key components: adaptive subspace projection for multi-source interest fusion, logarithmic time-decay factors for popularity bias mitigation, and dynamic gating mechanisms for personalized recommendation weighting. The framework uniquely combines sequential behavior analysis, social graph propagation, and temporal popularity modeling through a unified architecture. Experimental results on the MIND dataset, an open-source version of MSN News, demonstrate that PAD-MPFN outperforms existing methods in terms of recommendation performance and cold-start scenarios while effectively alleviating information overload. This study offers a new solution for dynamic interest modeling and diverse recommendation. Full article

Background/Objectives: Total thymectomy is currently the gold standard operation for treating thymoma. However, recent studies have suggested the potential health consequences of thymus removal in adults, including possible increased autoimmune disease and all-cause mortality. In this context, we assess oncologic outcomes following total vs. partial thymectomy for early-stage thymoma. Methods: We identified patients diagnosed with WHO types A–B3 thymoma between 2010–2021 from a national hospital-based dataset. We excluded patients with stage II or higher disease, open resections and perioperative chemo-/radiation therapy. We stratified patients into total and partial thymectomy cohorts. We used propensity score matching to minimize confounding, Kaplan–Meier analysis to estimate survival, and Cox proportional hazards to identify associations. Results: Of 1598 patients with early-stage thymoma, 495 (31.0%) underwent partial and 1103 (69.0%) total thymectomy. Patients undergoing partial thymectomy were similar in sex (female 53.7% vs. 53.4%; p = 0.914), race (white 74.5% vs. 74.0%; p = 0.921), comorbidities (0 in 77.0% vs. 75.5%; p = 0.742), and tumor size (48.7 mm vs. 50.4 mm; p = 0.455) compared to total thymectomy. There were no differences in 30-day (0.8% vs. 0.6%, p = 0.747) or 90-day mortality (0.8% vs. 0.8%, p > 0.999), which persisted after matching. Moreover, 10-year survival was similar in both unmatched (p = 0.471) and matched cohorts (p = 0.828). Partial thymectomy was not independently associated with survival (aHR = 1.00, p = 0.976). Conclusions: In patients with early-stage thymoma, partial and total thymectomy were associated with similar short- and long-term outcomes. In light of recent attention to the role of the thymus gland, the results add important insights to shared decision-making discussions. Full article

Semantic change detection technology based on remote sensing data holds significant importance for urban and rural planning decisions and the monitoring of ground objects. However, simple convolutional networks are limited by the receptive field, cannot fully capture detailed semantic information, and cannot effectively perceive subtle changes and constrain edge information. Therefore, a dynamic graph reasoning network with layer-by-layer semantic decomposition for semantic change detection in remote sensing data is developed in response to these limitations. This network aims to understand and perceive subtle changes in the semantic content of remote sensing data from the image pixel level. On the one hand, low-level semantic information and cross-scale spatial local feature details are obtained by dividing subspaces and decomposing convolutional layers with significant kernel expansion. Semantic selection aggregation is used to enhance the characterization of global and contextual semantics. Meanwhile, the initial multi-scale local spatial semantics are screened and re-aggregated to improve the characterization of significant features. On the other hand, at the encoding stage, the weight-sharing approach is employed to align the positions of ground objects in the change area and generate more comprehensive encoding information. Meanwhile, the dynamic graph reasoning module is used to decode the encoded semantics layer by layer to investigate the hidden associations between pixels in the neighborhood. In addition, the edge constraint module is used to constrain boundary pixels and reduce semantic ambiguity. The weighted loss function supervises and optimizes each module separately to enable the network to acquire the optimal feature representation. Finally, experimental results on three open-source datasets, such as SECOND, HIUSD, and Landsat-SCD, show that the proposed method achieves good performance, with an SCD score reaching 35.65%, 98.33%, and 67.29%, respectively. Full article

Intelligent monitoring and fault diagnosis of PV grid-connected inverters are crucial for the operation and maintenance of PV power plants. However, due to the significant influence of weather conditions on the operating status of PV inverters, the accuracy of traditional fault diagnosis methods faces challenges. To address the issue of open-circuit faults in power switching devices, this paper proposes a multi-dimensional feature perception network. This network captures multi-scale fault features under complex operating conditions through a multi-dimensional dilated convolution feature enhancement module and extracts non-causal relationships under different conditions using convolutional feature fusion with a Transformer. Experimental results show that the proposed network achieves fault diagnosis accuracies of 97.3% and 96.55% on the inverter dataset and the generalization performance dataset, respectively. Full article

Adversarial machine learning exploits the vulnerabilities of artificial intelligence (AI) models by inducing malicious distortion in input data. Starting with the effect of adversarial methods on well-known MNIST and CIFAR-10 open datasets, this paper investigates the ability of Uniform Manifold Approximation and Projection (UMAP) in providing useful representations of both legitimate and malicious images and analyzes the attacks’ behavior under various conditions. By enabling the extraction of decision rules and the ranking of important features from classifiers such as decision trees, eXplainable AI (XAI) achieves zero false positives and negatives in detection through very simple if-then rules over UMAP variables. Several examples are reported in order to highlight attacks behaviour. The data availability statement details all code and data which is publicly available to offer support to reproducibility. Full article

Urban environments present substantial obstacles to GPS positioning accuracy, primarily due to multipath interference and limited satellite visibility. To address these challenges, we propose a novel weighting approach, referred to as the HK model, that enhances real-time GPS positioning performance by leveraging the variability of the signal-to-noise ratio (SNR), without requiring auxiliary sensors. Analysis of 24 h observational datasets collected across diverse environments, including open-sky (OS), city streets (CS), and urban canyons (UC), demonstrates that multipath-affected non-line-of-sight (NLOS) signals exhibit significantly greater SNR variability than direct line-of-sight (LOS) signals. The HK model classifies received signals based on the standard deviation of their SNR and assigns corresponding weights during position estimation. Comparative performance evaluation indicates that relative to existing weighting models, the HK model improves 3D positioning accuracy by up to 22.4 m in urban canyon scenarios, reducing horizontal RMSE from 13.0 m to 4.7 m and vertical RMSE from 19.5 m to 6.9 m. In city street environments, horizontal RMSE is reduced from 11.6 m to 3.8 m. Furthermore, a time-sequential analysis at the TEHE site confirms consistent improvements in vertical positioning accuracy across all 24-hourly datasets, and in terms of horizontal accuracy, in 22 out of 24 cases. These results demonstrate that the HK model substantially surpasses conventional SNR- or elevation-based weighting techniques, particularly under severe multipath conditions frequently encountered in dense urban settings. Full article