Intelligent Systems with Applications最新文献

{"title":"Transformers for sign language processing: A survey of spatio-temporal modeling approaches, challenges, and research gaps","authors":"Thapelo Sindane,&nbsp;Vukosi Marivate,&nbsp;Abiodun Modupe,&nbsp;Herkulaas Combrink","doi":"10.1016/j.iswa.2026.200723","DOIUrl":"10.1016/j.iswa.2026.200723","url":null,"abstract":"<div><div>Transformer models have recently become the dominant paradigm for learning representations of complex sequential data. However, their behavior under extreme spatio-temporal structure, multimodality, and data scarcity remains insufficiently understood. This paper presents a structured and critical review of transformer-based approaches for sign language processing, using sign language as a stress test for transformer learning in low-resource, multi-modal, and spatially structured settings. We synthesize and analyze existing work in sign language translation and production, with particular emphasis on sign language production as a generative learning problem that exposes limitations in current modeling assumptions. Our review examines how transformer architectures originally developed for natural language processing and computer vision are adapted to capture spatial–temporal correlations, multi-stream inputs, and cross-modal alignment. We further analyze architectural choices, loss formulations, and evaluation protocols, identifying recurring design patterns as well as systematic shortcomings. Based on this analysis, we highlight open challenges and research opportunities for transformer-based learning in highly structured multimodal domains, with implications extending beyond sign language to other low-resource spatio-temporal learning problems.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"31 ","pages":"Article 200723"},"PeriodicalIF":5.5,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148950070","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"ArabLJP: an evidence-grounded and retrieval-augmented LLM framework for reliable Arabic legal judgment prediction in statute-centric legal systems","authors":"Ahmed Mohamed,&nbsp;Ezieddin Elmahjub,&nbsp;Khaled Shaban","doi":"10.1016/j.iswa.2026.200727","DOIUrl":"10.1016/j.iswa.2026.200727","url":null,"abstract":"<div><div>Legal judgment prediction (LJP) using large language models (LLMs) is increasingly explored; however, its practical deployment remains constrained by hallucinations and limited evidentiary traceability, challenges that are especially acute in under-resourced, statute-centric Arabic legal systems. This paper introduces ArabLJP, an evidence-grounded and retrieval-augmented framework that reframes Arabic LJP as statute-conditioned judgment generation rather than label-only classification. ArabLJP integrates (i) a reasoning-preserving Saudi commercial-law dataset that explicitly retains factual narratives, judicial reasoning, and operative rulings; (ii) a hybrid \"SMART RAG\" retrieval pipeline that combines lexical matching, semantic embeddings, and LLM-guided query refinement and reranking, with priority grounding in statutes explicitly cited in case materials; and (iii) a hallucination-aware, multi-dimensional evaluation protocol that measures semantic similarity, entity and numeric consistency, and outcome fidelity, with explicit penalties for unsupported factual inventions.</div><div>The framework is evaluated under controlled experimental conditions across three modeling paradigms (prompt-only generation, retrieval-augmented structured prompting, and parameter-efficient supervised adaptation) and multiple LLM backbones, including GPT-4o-mini and four open-weight models (LLaMA-3.2, Qwen-3, Phi-4, and DeepSeek). Empirical results demonstrate that evidence-grounded retrieval and structured prompting substantially reduce hallucination strength and improve ruling fidelity relative to prompt-only baselines, while supervised adaptation yields further gains in semantic and stylistic alignment with reference judgments.</div><div>Overall, ArabLJP shows that disciplined statutory grounding, structured judicial prompting, and reliability-oriented evaluation measurably improve LJP robustness in Arabic statute-centric legal settings. The framework provides a reproducible methodological baseline for studying and mitigating hallucinations in Arabic LJP, while underscoring that fully reliable automated legal reasoning remains an open research challenge.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"31 ","pages":"Article 200727"},"PeriodicalIF":5.5,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148950073","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"EDGE-YOLO11: A lightweight dual-stream RGB-T pedestrian detector under varying illumination","authors":"Jin Deng,&nbsp;Zhongmin Li,&nbsp;Yufeng Yao","doi":"10.1016/j.iswa.2026.200728","DOIUrl":"10.1016/j.iswa.2026.200728","url":null,"abstract":"<div><div>RGB-T pedestrian detection has attracted increasing attention because visible and infrared modalities provide complementary information under challenging illumination conditions. However, many existing RGB-T detectors rely on relatively heavy dual-stream or fusion structures, which limits their applicability in resource-constrained deployment scenarios. To address this issue, this paper proposes EDGE-YOLO11, a lightweight dual-stream RGB-T pedestrian detector under varying illumination.</div><div>The proposed framework processes visible and infrared inputs with a compact dual-stream architecture and selectively introduces GhostConv and DualConv into specific layers rather than replacing all standard convolutions. This design reduces redundant computation while preserving the mid-level RGB-T fusion pipeline. In addition, an efficient multi-scale attention module is employed as a lightweight feature recalibration mechanism to partially compensate for the representation loss introduced by lightweight convolutional compression.</div><div>Experiments are conducted on the LLVIP dataset with full validation and illumination-specific evaluation, including day, transition, and night subsets. Additional experiments are performed on a constructed FLIR-person subset to evaluate the applicability of EDGE-YOLO11 beyond LLVIP. Compared with YOLOv11-RGBT-s on LLVIP, EDGE-YOLO11-s achieves competitive detection performance with 96.31% mAP50 and 64.68% mAP50-95, while reducing parameters by 77.3% and FLOPs by 71.7%. In CPU ONNX Runtime inference, EDGE-YOLO11-s achieves 44.61 ms latency and 22.42 FPS, which is approximately 2.56<span><math><mo>×</mo></math></span> faster than YOLOv11-RGBT-s.</div><div>These results show that EDGE-YOLO11 provides a favorable efficiency–accuracy trade-off for deployment-oriented RGB-T pedestrian detection under varying illumination.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"31 ","pages":"Article 200728"},"PeriodicalIF":5.5,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148994891","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Explainable artificial intelligence: An expanded taxonomy by mechanism and interpretability perspectives","authors":"Abdelrahman Elsayed,&nbsp;Yuandi Wu,&nbsp;John Yawney,&nbsp;Stephen Andrew Gadsden","doi":"10.1016/j.iswa.2026.200709","DOIUrl":"10.1016/j.iswa.2026.200709","url":null,"abstract":"<div><div>Explainable Artificial Intelligence (XAI) has emerged as a critical area of research to address the opacity of modern machine learning models, with deep neural networks being a major focus. While numerous surveys of XAI exist, many either focus narrowly on specific techniques or lack a unifying structure that connects diverse approaches into a coherent framework. This paper presents a comprehensive literature review that consolidates both widely adopted and less commonly discussed XAI methods, including perturbation and gradient based explainers, dimensionality reduction techniques, counterfactual frameworks, and attribution methods. We adopt the well established taxonomy in literature, which categorizes XAI approaches into scoop based, complexity based, model based, and methodology based classes, and extend it in two important ways: (i) by introducing finer operational subcategories (e.g., gradient vs. perturbation based, model specific vs. model agnostic), and (ii) by introducing a complementary interpretability level distinction that categorizes explanations as structural, semantic, or behavioral according to the level at which understanding is achieved. For each method, we summarize its underlying principles and highlight reported strengths and limitations, enabling readers to understand both the technical foundations and the trade offs involved. By structuring the field into this extended taxonomy, the survey provides a road map for new researchers entering XAI and offers a reference point for identifying gaps where novel approaches can be developed.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"31 ","pages":"Article 200709"},"PeriodicalIF":5.5,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148950067","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Ameliorating tram service quality using a multi-attribute decision-making framework in a bipolar complex fuzzy environment","authors":"Abrar Hussain,&nbsp;Sarbast Moslem,&nbsp;Tapan Senapati,&nbsp;Domokos Esztergar-Kiss","doi":"10.1016/j.iswa.2026.200721","DOIUrl":"10.1016/j.iswa.2026.200721","url":null,"abstract":"<div><div>Service quality in tram systems plays a vital role in enhancing urban mobility, passenger satisfaction, and the overall efficiency of public transport networks. However, evaluating tram service quality involves handling vague, redundant, and sometimes conflicting human judgments. To address these challenges, this study proposes a novel multi-attribute decision-making (MADM) framework based on bipolar complex fuzzy (BCF) Sugeno–Weber aggregation operators, applied for the first time to a real-life tram service quality problem in Budapest. Some feasible Sugeno–Weber algebraic operations are formulated to integrate uncertainty effectively, incorporating the BCF environments. A family of innovative mathematical approaches is also developed using Sugeno-Weber operations, including the BCF Sugeno-Weber weighted average and geometric operators. An intelligent decision-making framework is established to evaluate ranking among different alternatives under multiple conflicting criteria and key features. To validate and demonstrate the compatibility of the proposed methodologies, an illustrative numerical example assesses five tram improvement alternatives across four key criteria. The effectiveness and reliability are verified through comprehensive sensitivity analysis and detailed comparison with existing approaches. Finally, the advantages and concluding remarks, along with future directions, are discussed.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"31 ","pages":"Article 200721"},"PeriodicalIF":5.5,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148950071","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhanced MPA for solving robot gripper and engineering design problems","authors":"Hussam Fakhouri ,&nbsp;Faten Hamad ,&nbsp;Niveen Halalsheh","doi":"10.1016/j.iswa.2026.200657","DOIUrl":"10.1016/j.iswa.2026.200657","url":null,"abstract":"<div><div>This paper proposes a novel Enhanced Marine Predators Algorithm (EMPA) that integrates several innovative enhancements to improve original MPA algorithm performance to solve complex optimization problems such as engineering design problems. EMPA introduces dynamic parameter control, dynamic population size adjustment, and the elite pool to EMPA. These enhancements enable EMPA to better balance exploration and exploitation, leading to superior optimization results. The algorithm’s performance is rigorously tested using benchmark functions from the IEEE CEC 2014, CEC 2017, and CEC 2022 competitions, where it consistently outperforms other state-of-the-art optimization algorithms. EMPA’s has been applied to various engineering design problems, including material minimization in cantilever beams and structural optimization in three-bar trusses, demonstrating its practical effectiveness and adaptability. The results demonstrate that the performance of EMPA with varying dimensions outperforms the compared optimizers. For instance, in experiments with a dimension equal to 10, EMPA achieved the best performance in 9 functions on the CEC2022 benchmark, 27 functions on the CEC2017 benchmark, and achieving the first or second-best optimizer ranking in 27 functions on the CEC2014 benchmark.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"30 ","pages":"Article 200657"},"PeriodicalIF":4.3,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147858198","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Retraction to “Sentiment Analysis: From Rule-Based Lexicons to Large Language Models” [Intelligent Systems with Applications 28 (2025) 200599]","authors":"Maikel Leon","doi":"10.1016/j.iswa.2026.200634","DOIUrl":"10.1016/j.iswa.2026.200634","url":null,"abstract":"","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"30 ","pages":"Article 200634"},"PeriodicalIF":4.3,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147956865","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Intelligent optimisation for onshore wind farm battery energy storage systems","authors":"Mandisi Gwabavu ,&nbsp;R.C. Bansal ,&nbsp;P.A. Owolawi","doi":"10.1016/j.iswa.2026.200667","DOIUrl":"10.1016/j.iswa.2026.200667","url":null,"abstract":"<div><div>Intelligent optimisation is essential for enhancing the performance and stability of onshore wind farms integrated with Battery Energy Storage Systems (BESSs), optimising turbine positioning, energy storage, and output to ensure efficient operation and sustained grid stability. Notwithstanding considerable advances in intelligent optimisation, there has been limited success in effectively integrating the diverse impacts of wind conditions and BESS state of health (SOH) while ensuring grid stability. This study investigates the application of intelligent optimisation for integrating BESS with onshore wind farms to augment energy storage capacity and ensure grid reliability. The proposed model incorporates a neural network (NN)-based inverse model, Bayesian optimisation, Gaussian Process Regression (GPR), and Reinforcement Particle Swarm Optimisation (RPSO) to enhance wind energy production while providing accurate estimates of system performance. Significant results from a case study of an onshore wind farm in South Africa and a 138 MW BESS demonstrate that the proposed intelligent optimisation model integrating NN-based inverse modelling, Bayesian-optimised GPR for SOH estimation, and RPSO enhance overall wind farm efficiency by 15-20% and reduce energy storage management errors by up to 10%. The proposed model is validated using NASA’s battery degradation datasets, achieving an RMSE of less than 1% in SOH estimation. These results highlight the potential of intelligent optimisation to enhance wind farms' performance, reliability, and stability integrated with battery energy storage systems, offering a scalable solution for renewable energy management.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"30 ","pages":"Article 200667"},"PeriodicalIF":4.3,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147802270","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Assessment of smile intensity using an ordinal scale toward QOL monitoring","authors":"Kei Shimonishi ,&nbsp;Junyao Zhang ,&nbsp;Kazuaki Kondo ,&nbsp;Hirotada Ueda ,&nbsp;Yuichi Nakamura","doi":"10.1016/j.iswa.2026.200668","DOIUrl":"10.1016/j.iswa.2026.200668","url":null,"abstract":"<div><div>Facial expressions are the most common resource for estimating the internal state of a person and are widely employed as a method for monitoring in care settings. In daily life, peak or high-intensity expressions appear for only short durations of time, whereas neutral or low-to-mid-intensity expressions are far more common. Existing algorithms for facial expression recognition (FER) have focused on detecting high-intensity expressions, which limits the detection of low-to-mid expressions and prevents detailed intensity assessments. To address this problem, we propose a framework for facial expression assessment based on comparison and ordinal scale. Using this framework, we also propose methods for determining the number of ordinal ranks based on discriminability and for selecting reference images for intensity assessment. Experimental results demonstrate that the proposed framework outperforms both human annotators and conventional FER systems.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"30 ","pages":"Article 200668"},"PeriodicalIF":4.3,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147802271","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Retraction Notice to “AI Safety Practices and Public Perception: Historical Analysis, Survey Insights, and a Weighted Scoring Framework” [Intelligent Systems with Applications 28 (2025) 200583]","authors":"Maikel Leon","doi":"10.1016/j.iswa.2026.200633","DOIUrl":"10.1016/j.iswa.2026.200633","url":null,"abstract":"","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"30 ","pages":"Article 200633"},"PeriodicalIF":4.3,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147956866","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}