Artificial Intelligence Review最新文献

{"title":"Methodological challenges in explainable AI for fraud detection: a systematic literature review","authors":"Usama Zafar,&nbsp;Fan Wu","doi":"10.1007/s10462-026-11516-7","DOIUrl":"10.1007/s10462-026-11516-7","url":null,"abstract":"<div><p>As complex black-box Artificial Intelligence (AI) models become integral to high-stakes domains like fraud detection, the need for transparency is becoming critical. Explainable AI (XAI) aims to make algorithmic decisions understandable to stakeholders. While systematic reviews have mapped the use of XAI in the broader financial sector, a focused synthesis of the methodological challenges unique to fraud detection remains a critical gap. This systematic literature review addresses this gap, following PRISMA 2020 guidance and synthesizing 49 peer-reviewed studies (2021–2025) to identify current practices and foundational challenges. Our analysis reveals that while post-hoc methods like SHAP and LIME are prevalent, their application is undermined by two systemic methodological flaws that threaten the validity of current research. First, we identify an explainability–imbalance paradox, where common data resampling techniques used to manage class imbalance can distort background distributions or local neighborhoods and thereby compromise the faithfulness and fidelity of post-hoc explanations. Second, we discover a profound evaluation vacuum, with around 80% of the analyzed studies using model predictive performance as a proxy for explanation quality rather than directly evaluating the explanations themselves. We also critically assess when post-hoc explanations are appropriate versus when intrinsically interpretable models are preferable in auditable and high-risk contexts. Based on these findings, we propose an explanation-aware training and evaluation checklist and a research agenda to guide the field toward standardized, explanation-level evaluation and the development of imbalance-robust, explanation-preserving data-processing methods.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 4","pages":""},"PeriodicalIF":13.9,"publicationDate":"2026-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11516-7.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147441651","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A systematic review of Green Machine Learning: practices and challenges for sustainability","authors":"Samara Santos,&nbsp;André L. C. Ottoni,&nbsp;Rita Borgo,&nbsp;Danton Ferreira,&nbsp;Erivelton Nepomuceno","doi":"10.1007/s10462-026-11515-8","DOIUrl":"10.1007/s10462-026-11515-8","url":null,"abstract":"<div><p>The growing energy demands and environmental impact of Machine Learning (ML) models have established Green Machine Learning (GML) as an important field for sustainable AI progress. This systematic review incorporates recent advances in energy-aware ML, contributing an overall assessment of algorithmic strategies, energy-aware design, and sustainability practices that reduce carbon footprints while preserving performance. The review identifies five basic dimensions of GML: algorithmic efficiency, energy optimization, hardware efficiency, carbon footprint tracking, and green benchmarks. Key findings indicate that pruning, quantization, and low-precision training can reduce energy consumption without significant loss of accuracy. Energy-aware federated learning and lightweight architectures have demonstrated potential savings in distributed environments. In contrast to prior work, this review integrates a full ML life-cycle perspective and contextualizes technical methods within broader sustainability frameworks, including the UN Sustainable Development Goals. It also surveys emerging trends, including energy-aware Federated Learning, Automated ML under <span>(text{CO}_2)</span> constraints, and low-power inference on edge devices. Critical challenges remain, including the lack of standardized reporting metrics, barriers to reproducibility, and the trade-off between energy savings and predictive accuracy. This review offers structured research opportunities to support the development of GML across various domains. It serves as a reference for researchers and practitioners committed to integrating environmental responsibility into ML innovation.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 5","pages":""},"PeriodicalIF":13.9,"publicationDate":"2026-02-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11515-8.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147642978","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"LegalAI Research in LLM Era: Data, Modeling and Evaluation","authors":"Xiao Chi,&nbsp;Wei Wang,&nbsp;Ziyao Zhang,&nbsp;Ang Li,&nbsp;Yuting Huang,&nbsp;Yiquan Wu,&nbsp;Kun Kuang,&nbsp;Changlong Sun,&nbsp;Xiaozhong Liu,&nbsp;Fei Wu,&nbsp;Minghui Xiong","doi":"10.1007/s10462-026-11514-9","DOIUrl":"10.1007/s10462-026-11514-9","url":null,"abstract":"<div><p>Legal artificial intelligence (LegalAI) refers to the use of artificial intelligence technologies to automate various legal tasks. Recent advances in large-scale language models have significantly enhanced the capabilities of LegalAI, marking a new stage in its development. In this paper, we present a comprehensive survey of how large language models (LLMs) are reshaping the research paradigm of LegalAI. Beyond improving task performance, LLMs now serve as integral components across the perspectives of data, modeling, and evaluation. We propose a role-based schema that categorizes the involvement of LLMs along these perspectives and use it to systematically review existing studies in three major legal tasks, including legal classification, legal retrieval, and legal generation. Besides, we conduct a detailed quantitative comparison of LLM effectiveness across roles and tasks, and our findings reveal that the impact of LLMs is shaped by both their assigned roles and the nature of the legal tasks.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 4","pages":""},"PeriodicalIF":13.9,"publicationDate":"2026-02-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11514-9.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147441723","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep learning based visually rich document content understanding: a survey","authors":"Yihao Ding,&nbsp;Soyeon Caren Han,&nbsp;Jean Lee,&nbsp;Eduard Hovy","doi":"10.1007/s10462-025-11477-3","DOIUrl":"10.1007/s10462-025-11477-3","url":null,"abstract":"<div><p>Visually Rich Documents (VRDs) play a vital role in domains such as academia, finance, healthcare, and marketing, as they convey information through a combination of text, layout, and visual elements. Traditional approaches to extracting information from VRDs rely heavily on expert knowledge and manual annotation, making them labor-intensive and inefficient. Recent advances in deep learning have transformed this landscape by enabling multimodal models that integrate vision, language, and layout features through pretraining, significantly improving information extraction performance. This survey presents a comprehensive overview of deep learning-based frameworks for VRD Content Understanding. We categorize existing methods based on their modeling strategies and downstream tasks, and provide a comparative analysis of key components, including feature representation, fusion techniques, model architectures, and pretraining objectives. Additionally, we highlight the strengths and limitations of each approach and discuss their suitability for different applications. The paper concludes with a discussion of current challenges and emerging trends, offering guidance for future research and practical deployment in real-world scenarios.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 4","pages":""},"PeriodicalIF":13.9,"publicationDate":"2026-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-025-11477-3.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147441603","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Geospatial reasoning and awareness in large language models: a systematic review","authors":"Gabriel Ionut Dorobantu,&nbsp;Ana Cornelia Badea","doi":"10.1007/s10462-026-11512-x","DOIUrl":"10.1007/s10462-026-11512-x","url":null,"abstract":"<div><p>This research examines the evolution and integration of large language models within the geospatial domain, exploring both theoretical aspects and practical applications. Using a systematic review guided by the PRISMA methodology, the study investigates GeoAI developments and assesses the impact of foundation models in geospatial contexts. The findings highlight that commercial models demonstrate notable capabilities in interpreting geospatial concepts and generating functional code, although they face limitations concerning accessibility, transparency and reliance on external infrastructures. Smaller, open-source models, adapted through approaches such as fine-tuning and Retrieval-Augmented Generation, are identified as feasible alternatives, providing a balanced solution in terms of accuracy, efficiency and customization. The study emphasizes a need for large-scale, standardized datasets for effective training and evaluation of geospatial models, pointing toward a clear direction for future research. Despite significant advancements, achieving full autonomy of geospatial agents in complex task-solving scenarios remains an unresolved challenge. The future progression of GeoAI will rely heavily on interdisciplinary collaboration and the development of robust, transparent and ethical models capable of supporting real-time decision-making and promoting digital transformation in public administration and related fields.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 4","pages":""},"PeriodicalIF":13.9,"publicationDate":"2026-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11512-x.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147338422","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Semantic and temporal-aware hybrid embedding for transformer-based sequential recommendation","authors":"Smriti Vipin Madangarli,&nbsp;G. Suganeshwari,&nbsp;S. P. Syed Ibrahim,&nbsp;Muthu Subash Kavitha,&nbsp;Norio Setozaki","doi":"10.1007/s10462-025-11483-5","DOIUrl":"10.1007/s10462-025-11483-5","url":null,"abstract":"<div><p>Sequential recommendation has achieved remarkable progress with self-attentive architectures such as SASRec (Self-Attentive Sequential Recommendation), yet existing approaches often underutilize semantic relationships among items and fail to model temporal dynamics effectively. These limitations reduce the ability of existing systems to capture nuanced user preferences in real-world settings. To address these issues, this work introduces two novel frameworks: one for effective semantic integration and another for a semantic and temporal-aware hybrid embedding generation that enhances the representational capacity of SASRec. These frameworks construct a semantically rich hybrid matrix utilizing <b>M</b>arkov State Transition probabilities, <b>C</b>osine similarity, <b>P</b>ersonalized PageRank(PPR) based normalization and additionally <b>T</b>ime-weighted decay information for the temporal variant. The constructed hybrid matrix is smoothed using a graph convolutional network (GCN) to generate item embeddings, which are then passed to the transformer-based sequential recommendation model SASRec for next-item prediction. Evaluated on three real-world benchmark datasets (MovieLens, Yelp and Amazon Beauty), our proposed temporal variant achieves up to 10.4% improvement in HR@10 and 8.2% in NDCG@10 over SASRec, while maintaining competitive efficiency. Our studies confirm the effectiveness of each component, including semantic graph construction, temporal weighting, and contrastive alignment. These results demonstrate that incorporating semantic and temporal signals into sequential recommenders substantially enhances both accuracy and relevancy of recommendations.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 4","pages":""},"PeriodicalIF":13.9,"publicationDate":"2026-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-025-11483-5.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147338421","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Glider snake optimizer (GSO): a nature-inspired metaheuristic algorithm for global and engineering optimization problems","authors":"El-Sayed M. El-kenawy,&nbsp;Nima Khodadadi,&nbsp;Seyedali Mirjalili,&nbsp;Ahmed Mohamed Zaki,&nbsp;Abdelhameed Ibrahim,&nbsp;Amel Ali Alhussan,&nbsp;Doaa Sami Khafaga,&nbsp;Marwa M. Eid","doi":"10.1007/s10462-026-11504-x","DOIUrl":"10.1007/s10462-026-11504-x","url":null,"abstract":"<div><p>The rapid expansion of complex engineering and real-world optimization problems necessitates the development of efficient, adaptable, and computationally lightweight metaheuristic algorithms. In this study, a novel nature-inspired algorithm called glider snake optimization (GSO) is proposed, which draws behavioral inspiration from the gliding and serpentine locomotion patterns of arboreal snakes to enhance solution exploration and convergence control. The GSO algorithm incorporates a multi-segment movement mechanism, a flexible gliding path generator, and an elite guidance model to ensure effective balance between exploration and exploitation. Extensive experimental validation is conducted using a comprehensive set of 23 classical benchmark functions, high-dimensional test cases (100D, and 500D), the CEC 2019 benchmark suite, and several constrained engineering design problems. The results demonstrate that GSO outperforms or matches 13 state-of-the-art algorithms, including particle swarm optimization (PSO), grey wolf optimizer (GWO), whale optimization algorithm (WOA), and differential evolution (DE) in terms of accuracy, convergence speed, computational cost, and robustness. The algorithm also exhibits exceptional stability across parameter variations, as confirmed through sensitivity analysis and statistical significance testing. These findings highlight the potential of GSO as a powerful and efficient tool for solving complex optimization problems in both theoretical and practical domains. Additionally, GSO achieves leading performance on most benchmark functions, with error reductions of up to 90% compared with competing algorithms. GSO also demonstrates faster convergence and lower variance across repeated trials, confirming its robustness. These quantitative outcomes further reinforce the effectiveness of the proposed algorithm. The MATLAB and Python implementations of GSO are available at https://nimakhodadadi.com.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 3","pages":""},"PeriodicalIF":13.9,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11504-x.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147338271","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Quality over quantity: a data-centric survey of annotation errors in object detection datasets","authors":"Adnan Hussain,&nbsp;Kaleem Ullah,&nbsp;Muhammad Afaq,&nbsp;Muhammad Munsif,&nbsp;Altaf Hussain,&nbsp;Sung Wook Baik","doi":"10.1007/s10462-026-11502-z","DOIUrl":"10.1007/s10462-026-11502-z","url":null,"abstract":"<div><p>In recent years, object detection has become a cornerstone of many computer vision applications, relying heavily on the availability of high-quality annotated datasets. However, even widely used benchmarks often suffer from annotation issues such as inaccurate bounding boxes, misclassified objects, and missing labels. These annotation errors, especially localization errors, can greatly affect the training and evaluation of detection models. In this survey, we provide a data-centric and comprehensive review of existing methods for identifying and analyzing errors in object detection datasets. We examine the main components of error detection workflows, including annotation error taxonomies and model-agnostic detection techniques. In addition, we develop a standardized categorization of annotation error types specific to object detection, providing a foundation for consistent analysis and comparison across studies. We also perform manual inspections of selected benchmark datasets to observe and quantify common annotation errors in practice. Moreover, the survey highlights the datasets used for evaluating error detection methods and compares their scope and inherent challenges. Finally, we summarize the types of annotation errors found in existing benchmarks and provide recommendations for future research to enhance dataset quality and reliability in object detection.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 3","pages":""},"PeriodicalIF":13.9,"publicationDate":"2026-02-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11502-z.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147337682","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Harnessing deep learning for air pollution forecasting: trends, techniques, and future prospects","authors":"Salvator Lawrence,&nbsp;Srimuruganandam Bhathmanabhan","doi":"10.1007/s10462-026-11496-8","DOIUrl":"10.1007/s10462-026-11496-8","url":null,"abstract":"<div><p>Air pollution is a serious global public health threat arising from exposure to toxic ambient pollutants, including particulate matter (PM), sulphur oxides (SOx), nitrogen oxides (NOx), ozone (O₃), carbon monoxide (CO), and ammonia (NH₃). Traditional statistical and deterministic forecasting models often fail to adequately represent nonlinear interactions among multiple pollutants, meteorological drivers, and anthropogenic influences, motivating the growing adoption of deep learning (DL) approaches. This systematic review synthesizes evidence from more than 150 peer-reviewed studies conducted across diverse geographical regions and employing a wide range of DL architectures, including standalone, hybrid, and advanced spatiotemporal models. Using structured quantitative summaries, rank-based performance comparisons, and methodological assessments, the review identifies leading model families, analyzes pollutant- and horizon-specific performance trends, and evaluates robustness and generalizability across spatial and temporal contexts. Overall, DL models generally outperform traditional approaches, particularly when multi-source inputs and spatiotemporal dependencies are explicitly modeled. Nevertheless, the literature remains fragmented, with a strong concentration of studies in data-rich urban regions of Asia, heterogeneous datasets, inconsistent evaluation protocols, limited transparency, and weak external validity. Addressing these limitations requires standardized preprocessing and benchmarking practices, improved explainability and uncertainty quantification, and the development of globally representative datasets. Emerging directions, including hybrid, physics-informed, and generative DL architectures, offer promising pathways to enhance reliability and operational deployment. Collectively, this review provides a comprehensive and critical synthesis of DL-based air quality forecasting, offering actionable insights for researchers, practitioners, and policymakers seeking transparent, generalizable, and policy-relevant prediction systems for environmental management and public health protection.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 3","pages":""},"PeriodicalIF":13.9,"publicationDate":"2026-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11496-8.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147337305","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dynamic causal modeling and efficient fuzzy temporal causality reasoning in industrial fault diagnosis: a C-DUCG approach","authors":"Chunling Dong,&nbsp;Jianxiang Cao,&nbsp;Wenjing Liu","doi":"10.1007/s10462-026-11509-6","DOIUrl":"10.1007/s10462-026-11509-6","url":null,"abstract":"<div>\u0000 \u0000 <p>Online fault diagnosis is crucial for improving the reliability of safety–critical industrial systems. Fault diagnosis becomes increasingly complex if the time-variations, fuzziness, and uncertain causal relationships related to the various internal factors in the system are present. In view of the time-varying working conditions of industrial systems, as well as the fuzzy uncertainty of fault data and knowledge, a C-DUCG (Cubic dynamic uncertain causality graph) approach is proposed for fault diagnosis. Such a solution is intended to help develop a generative cubic causality graph modeling scheme and a FUTURE (fuzzy temporal causality reasoning) algorithm, both of which facilitate the representation and reasoning about complex fault situations (involving temporal causalities and uncertain evidence). In C-DUCG, the strategies of causality simplification and EELA (event-oriented early logical absorption) are proposed to mitigate the complexities of modeling and reasoning. Comparative experiments and a sequence of fault diagnosis tests on a nuclear power plant (NPP) simulator validate the efficiency, recall, accuracy, and interpretability of C-DUCG in large-scale dynamic systems. Experiments reveal that the proposed algorithm achieves 0.62 and 0.999 in terms of recall rate AC@1 and AC@k, respectively, with the average reasoning time being 10 ms, and the average time spent in NPP fault diagnosis tests is 9.42 ms.</p>\u0000 </div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 3","pages":""},"PeriodicalIF":13.9,"publicationDate":"2026-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11509-6.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147337077","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}