Artificial Intelligence Review最新文献

{"title":"Deep learning for real-time strawberry detection, ripeness classification, and picking point localization: a review of architectures, field studies, and open challenges","authors":"Mohieddine Jelali,&nbsp;Fabian Gerz,&nbsp;Orhan-Timo Altan,&nbsp;Loui Al-Shrouf","doi":"10.1007/s10462-026-11747-8","DOIUrl":"10.1007/s10462-026-11747-8","url":null,"abstract":"<div><p>Accurate detection of strawberry fruit, reliable ripeness estimation, and precise localization of the picking point are essential for automated harvesting and yield prediction in smart farming. However, real-world environments introduce significant challenges, including occlusion, illumination variability, and high visual similarity between ripeness stages. Deep learning (DL)-based object detection methods have become the dominant approach to address these issues. This paper presents a systematic review of 60 peer-reviewed studies published between 2023 and 2026, focusing on detecting strawberries and their ripeness using the YOLO family (v5–v11) of DL algorithms. The studies are analyzed with respect to dataset characteristics, preprocessing and augmentation strategies, model architectures, and evaluation protocols. Our results show a clear dominance of YOLOv8 (28 studies, 46.7%) and YOLOv11 (13 studies, 21.7%), while hybrid CNN–ViT models are gaining attention (8 studies, 13.3%). However, critical methodological issues are identified: in 21 out of 60 studies (35.0%), data augmentation was applied before dataset splitting, 13 studies (21.7%) adhered to the recommended split-then-augment protocol, and 8 studies (13.3%) applied online augmentation. The remaining 18 studies (30.0%) did not report using any augmentation. Reproducibility is severely limited by scarce sharing of code or trained weights (2 studies, 3.3%) and datasets (8 studies, 13.3%). To quantify the empirical impact of this methodological flaw, we conducted a controlled experiment on the StrawDI_Db1 dataset using YOLOv8, comparing four augmentation strategies. The results demonstrate that the full leakage protocol overestimates test performance by 11.76 to 19.28 percentage points (p.p.) in mAP@0.5:0.95 (and by 8.02 to 11.50 p.p. in mAP@0.50) compared to a clean, split-then-augment pipeline. The underlying validation-test discrepancy reaches up to 18.63 p.p. (mAP@0.5:0.95) and 13.81 p.p. (mAP@0.50). These findings provide direct quantitative evidence that the performance metrics reported in 35.0% of the reviewed studies may be substantially inflated. This review provides a structured overview of current approaches, a critical assessment of existing research practices, and actionable guidance for developing robust, deployment-ready DL solutions for precision agriculture.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 12","pages":""},"PeriodicalIF":18.8,"publicationDate":"2026-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11747-8.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149339456","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":"Seeing ahead through a state-of-the-art review of visual perception and predictive intelligence for collaborative robots","authors":"Mahmoud Salem,&nbsp;Markus Reischl,&nbsp;Veit Hagenmeyer,&nbsp;Katja Nau,&nbsp;Ahmed Elkaseer","doi":"10.1007/s10462-026-11708-1","DOIUrl":"10.1007/s10462-026-11708-1","url":null,"abstract":"<div><p>The rate at which collaborative robots (cobots) are evolving from isolated, pre-programmed manipulators into collaborative partners that operate productively alongside humans, other robots, and machines is accelerating rapidly. Advances in computer vision is enabling cobots, even in dynamic and cluttered settings, to detect, recognise and assess both environment and actions. However, achieving higher detection accuracy and real-time localization in dynamic systems remains a challenge, as most existing approaches assume uncluttered backgrounds with limited scene complexity. Closing this gap is critical for both research and industrial deployment. This paper reviews the development of visual perception in cobots, surveys emerging computer vision algorithms, and examines a holistic perception pipeline in which raw vision signals, vision language models (VLMs), and augmented/virtual reality interfaces jointly power a unified sense-think-act loop. The review finds that cobots in high-mix, reconfigurable factories face serious challenges: occlusions in fields of view, fluctuating illumination, ambiguous backgrounds, few- or zero-shot object scenarios, high-speed motion, and computational problems in fusing heterogeneous data streams in real time. Cobots also need to manipulate damaged items, requiring vision algorithms that infer geometry from sparse cues and support adaptive grasping. Finally, the review identifies three complementary research directions to unlock plug-and-play cobots: (1) information aware sensor-fusion pipelines transmitting only credible voxels to reduce bandwidth while preserving sub-millimetre accuracy; (2) language conditioned planning and control networks mapping plain-language goals to smooth, jerk-bounded trajectories; and (3) pre-trained few-shot reward heads that score task progress without bespoke metrics, closing the reinforcement learning (RL) loop. Combined, these advances promise cobots that can be commissioned with relatively few demonstrations.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 12","pages":""},"PeriodicalIF":18.8,"publicationDate":"2026-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11708-1.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149234385","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":"Towards optimized forward–forward training for neural networks","authors":"Yomna Abdulgawad Elnady,&nbsp;Amany Mahmoud Sarhan,&nbsp;Mohammad Ali Eita","doi":"10.1007/s10462-026-11668-6","DOIUrl":"10.1007/s10462-026-11668-6","url":null,"abstract":"<div><p>This study investigates and enhances the Forward–Forward (FF) algorithm, a novel neural network training approach recently introduced by Geoffrey Hinton. Unlike Back-Propagation (BP), which relies on separate forward and backward passes, FF performs two forward passes, one with positive (real) data and one with negative data, potentially generated by the network itself. We evaluate FF across six diverse datasets to assess its general performance and employ two well-established and four state-of-the-art metaheuristic algorithms on four datasets to optimize its hyper-parameters. While FF offers a notable advantage in memory efficiency, making it suitable for low-resource hardware, it initially lags behind BP in predictive accuracy. Through automated hyper-parameter optimization, FF demonstrates substantial improvements, with validation accuracy increasing from 93% to 97% on Pneumonia-MedMnist, 87% to 91% on Fashion-MNIST, 82% to 94% on OrganC-MedMnist, and 88% to 95% on OrganA-MedMnist, while maintaining or reducing training time. These results demonstrate that metaheuristic optimization not only accelerates convergence but also significantly enhances FF’s predictive performance, bridging the gap between memory efficiency and accuracy.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 11","pages":""},"PeriodicalIF":18.8,"publicationDate":"2026-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11668-6.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149133376","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":"Unsupervised state-learning for multivariate time series data: a survey and practical guidelines","authors":"Phillip Johann Overlöper,&nbsp;Cecília Coelho,&nbsp;Oliver Niggemann","doi":"10.1007/s10462-026-11704-5","DOIUrl":"10.1007/s10462-026-11704-5","url":null,"abstract":"<div><p>Unsupervised state-learning aims to infer latent variables or representations from multivariate time series that can serve as predictive summaries of a system evolution. Depending on the modelling assumptions, these summaries may appear as discrete regimes, posterior belief states, continuous latent trajectories or learnt embeddings that are subsequently interpreted or regularised as states. State-learning enables segmentation, anomaly detection, predictive maintenance and clinical monitoring, but the literature is fragmented across probabilistic regime-switching models, clustering and change-point methods, as well as modern representation learning. This survey unifies these strands and provides a synthesis aimed at practical method selection. We formalise the notion of state and introduce a taxonomy that makes modelling choices explicit along key dimensions: discrete versus continuous states, observable versus latent representations, output formats (hard assignments, posteriors, embeddings) and fixed versus adaptive state cardinality. We then review three major methodological families: probabilistic models, clustering/segmentation approaches, as well as self-supervised or generative representation learning. We highlight core assumptions, computational trade-offs and failure modes under missingness, irregular sampling, non-stationarity, as well as scarce evaluation labels. Building on this synthesis, we present a method-selection framework and implementation-oriented workflows that summarise recurring patterns in the literature and our practical reading of them, covering preprocessing, model fitting, diagnostics and uncertainty handling. Furthermore, we propose an evaluation and reporting protocol tailored to state-learning, emphasising leakage-free temporal splits, temporal coherence, interpretability, complexity control and uncertainty calibration. Finally, we summarise widely used benchmark testbeds that provide evaluation-only labels.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 11","pages":""},"PeriodicalIF":18.8,"publicationDate":"2026-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11704-5.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149007447","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":"Quantum circuit-based learning models: bridging quantum computing and machine learning","authors":"Fan Fan,&nbsp;Yilei Shi,&nbsp;Mihai Datcu,&nbsp;Bertrand Le Saux,&nbsp;Luigi Iapichino,&nbsp;Francesca Bovolo,&nbsp;Silvia Liberata Ullo,&nbsp;Xiao Xiang Zhu","doi":"10.1007/s10462-026-11686-4","DOIUrl":"10.1007/s10462-026-11686-4","url":null,"abstract":"<div><p>Machine Learning (ML) has been widely applied across numerous domains due to its ability to automatically identify informative patterns from data for various tasks. The availability of large-scale data and advanced computational power enables the development of sophisticated models and training strategies, leading to state-of-the-art performance, but it also introduces substantial challenges. Quantum Computing (QC), which exploits quantum mechanisms for computation, has attracted growing attention and significant global investment as it may address these challenges. Consequently, Quantum Machine Learning (QML), the integration of these two fields, has received increasing interest, with a notable rise in related studies in recent years. We are motivated to review these existing contributions regarding quantum circuit-based learning models for classical data analysis and highlight the identified potentials and challenges of this technique. Specifically, we focus not only on QML models, both kernel-based and neural network-based, but also on recent explorations of their integration with classical machine learning layers within hybrid frameworks. Moreover, we examine both theoretical analysis and empirical findings to better understand their capabilities, and we also discuss the efforts on noise-resilient and hardware-efficient QML that could enhance its practicality under current hardware limitations. In addition, we cover several emerging paradigms for advanced quantum circuit design and highlight the adaptability of QML across representative application domains. This study aims to provide an overview of the contributions made to bridge quantum computing and machine learning, offering insights and guidance to support its future development and pave the way for broader adoption in the coming years.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 10","pages":""},"PeriodicalIF":18.8,"publicationDate":"2026-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11686-4.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148926829","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":"How metacognitive architectures remember their own thoughts: a systematic scoping review","authors":"Robin Nolte,&nbsp;Mihai Pomarlan,&nbsp;Ayden Janssen,&nbsp;Daniel Beßler,&nbsp;Kamyar Javanmardi,&nbsp;Sascha Jongebloed,&nbsp;Robert Porzel,&nbsp;John Bateman,&nbsp;Michael Beetz,&nbsp;Rainer Malaka","doi":"10.1007/s10462-026-11669-5","DOIUrl":"10.1007/s10462-026-11669-5","url":null,"abstract":"<div><p>Metacognition has gained significant attention for its potential to enhance autonomy and adaptability of artificial agents but remains a fragmented field: diverse theories, terminologies, and design choices have led to disjointed developments and limited comparability across systems. Existing overviews remain at a conceptual level that is undiscerning to the underlying algorithms, representations, and their respective effectiveness. We address this gap by performing a systematic scoping review. Reports were included if they described techniques enabling Computational Metacognitive Architectures (CMAs) to model, store, remember, and process their episodic <i>metacognitive experiences</i>, one of Flavell’s (1979) three foundational components of metacognition. Searches were conducted in 16 databases between December 2023 and June 2024. Data were charted using a 20-item framework considering pertinent aspects and analysed via structured narrative synthesis. A total of 101 reports on 35 distinct CMAs were included. Our findings show that metacognitive experiences may boost system performance and explainability, e.g., via self-repair. However, lack of standardisation and limited evaluations may hinder progress: only 17% of CMAs were quantitatively evaluated regarding this review’s focus, and significant terminological inconsistency limits cross-architecture synthesis. Systems also varied widely in memory content, data types, and employed algorithms. Limitations include the non-iterative nature of the search query, heterogeneous data availability, and an under-representation of sub-symbolic CMAs. Future research should focus on standardisation and evaluation, e.g., via community-driven challenges, and on transferring promising principles to emergent systems.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 10","pages":""},"PeriodicalIF":18.8,"publicationDate":"2026-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11669-5.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148927828","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":"Agi-powered affective computing in cyber-physical-social-thinking space: framework, techniques, applications and future directions","authors":"Feifei Shi,&nbsp;Jingyi Yang,&nbsp;Zita Lifelo,&nbsp;Jianguo Ding,&nbsp;Huansheng Ning","doi":"10.1007/s10462-026-11691-7","DOIUrl":"10.1007/s10462-026-11691-7","url":null,"abstract":"<div><p>The advent of Artificial General Intelligence (AGI) promises to revolutionize affective computing. However, current research lacks a systematic framework that accounts for the inherently cross-space nature of human emotion, where physiological states (physical space), cultural norms (social space), cognitive processes (thinking space), and digital data (cyberspace) continuously interact and shape each other. To address this gap, this paper starts by repositioning affective computing within the theory of Cyber-Physical-Social-Thinking (CPST) space, conceptualizing it as a dynamic process of cross-space transformation rather than an isolated function. AGI capabilities of perception, learning, reasoning, memory, generalization, and generation serve as the computational engines that enable these cross-space transformations. We then systematically map AGI techniques (e.g., causal modeling, chain-of-thought, diffusion models) to specific affective computing tasks, showing how each technique supports their realization. By integrating CPST as a theoretical backbone and AGI as an enabling technology, this paper provides a systematic roadmap for next-generation affective computing. Potential applications and future directions are also discussed, highlighting how this framework advances human-level affective intelligence toward genuine empathy and contextual awareness.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 10","pages":""},"PeriodicalIF":18.8,"publicationDate":"2026-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11691-7.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148927826","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":"Evidence of impact and interpretational limits of generative AI in STEM education: a systematic review and meta-analysis on cognitive learning outcomes","authors":"Chiara Boolzen,&nbsp;Jochen Kuhn,&nbsp;Salome Flegr,&nbsp;Eva-Maria Rott,&nbsp;Niklas Stausberg,&nbsp;Stefan Küchemann","doi":"10.1007/s10462-026-11665-9","DOIUrl":"10.1007/s10462-026-11665-9","url":null,"abstract":"&lt;div&gt;&lt;p&gt;This systematic review and meta-analysis examines the impact of generative artificial intelligence (GAI) on cognitive learning outcomes in STEM education. Prior research is growing but remains fragmented, often focusing on usability or single tools like ChatGPT rather than domain-specific cognitive effects. We therefore address these gaps by examining (1) the extent to which interactions with GAI enhance learning effectiveness and possible moderators, (2) what challenges learners face when interacting with GAI systems, and (3) which interventions support successful learner-GAI interaction. We meta-analyzed externally assessed cognitive outcomes (RQ1) and narratively synthesized reported learner challenges and supportive instructional interventions (RQ2-RQ3) when quantitative pooling was not feasible. Two pairs of raters independently screened and coded peer-reviewed quantitative studies published after 2017 that included a comparison/control group and examined cognitive learning in STEM involving learner-GAI interaction. A systematic search (ERIC, PsycINFO, Web of Science; updated May 7, 2026) and citation tracking yielded 85 eligible studies, of which 49 studies with 59 effect sizes met meta-analytic criteria. Most studies focused on higher education and used text-based GAI tools (e.g., ChatGPT). A random-effects meta-analysis shows an overall positive effects of GAI in STEM education, but the studies included exhibit a substantial heterogeneity (&lt;span&gt;(I^2=96.32%)&lt;/span&gt;), and the prediction interval ranges from Hedge’s &lt;span&gt;(g = -1.52)&lt;/span&gt; to &lt;span&gt;(g = 3.20)&lt;/span&gt;. In line with this, funnel plot asymmetry suggests a potential publication bias. To account for potential publication bias, we also conducted a Robust Bayesian Meta-Analysis (RoBMA) and found that the overall positive effect of GAI in STEM education can be largely attributed to publication bias (&lt;span&gt;(mu =0.076 pm 0.254)&lt;/span&gt;), however a large heterogeneity remains (&lt;span&gt;(tau =1.190 pm 0.166)&lt;/span&gt;), which appears to be not associated to publication bias. To explain the heterogeneity, we conducted a moderator analysis and found that the learning outcome (knowledge vs. skills) as well the &lt;i&gt;inversion-substitution-augmentation-redefinition&lt;/i&gt; (ISAR)-level, which compares the cognitive activity of the intervention and control in terms of the &lt;i&gt;interactive-constructive-active-passive&lt;/i&gt; (ICAP)-level, both explain parts of the observed heterogeneity (&lt;span&gt;(R^2=12.5 %)&lt;/span&gt; and &lt;span&gt;(R^2=13.0 %)&lt;/span&gt;, respectively). In contrast to the main effect, the moderator findings were robust under publication-bias correction using RoBMA. Apart from that, a coincidence analysis suggested that no combination of learning outcome type and ISAR level fulfills a sufficient condition for large effect sizes. Additionally, knowledge as a learning outcome type was found to be an almost, but not strictly, necessary condition for large effects when learning with generative AI. Ove","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 10","pages":""},"PeriodicalIF":18.8,"publicationDate":"2026-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11665-9.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148987125","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":"Epistemic norms for AI safety and alignment research","authors":"Keivan Navaie","doi":"10.1007/s10462-026-11666-8","DOIUrl":"10.1007/s10462-026-11666-8","url":null,"abstract":"<div><p>Mainstream AI research emphasises capability growth and tolerates low failure rates when average-case performance is high. AI safety and alignment research has a different mission: to ensure that catastrophic failures never occur, under sparse evidence, adversarial dynamics, and fat-tailed risk. We argue that the two domains differ along two analytically independent axes — <i>capability profile</i> (demonstrating the absence of hazardous behaviours versus the presence of positive capabilities) and <i>risk profile</i> (bounding worst-case outcomes under fat-tailed uncertainty versus optimising average-case performance) — and that mainstream epistemic practices are inadequate on both. Building on a structured synthesis grounded in a preregistered bibliometric baseline, we identify five cross-cutting gap dimensions in current alignment research, including the near-absence of institutionalised independent verification. To address these gaps we propose ECAISA, an Epistemic Code for AI Safety and Alignment comprising eight principles, a three-level scoring rubric, a four-level disclosure ladder that reconciles transparency with information-hazard and commercial-confidentiality constraints, a tiered applicability scheme, an infohazard adjudication procedure, and seven anti-gaming mechanisms. ECAISA does not certify that any AI system is safe; it constrains how safety-relevant research claims are documented, checked, and relied upon, with <i>auditability rather than certification</i> as its governance target. A retrospective rubric audit (κ = 0.79) demonstrates instrument feasibility; a four-stage validation roadmap is proposed.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 12","pages":""},"PeriodicalIF":18.8,"publicationDate":"2026-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11666-8.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149339040","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":"Embedding learning on multiplex networks for link prediction","authors":"Orell Trautmann,&nbsp;Olaf Wolkenhauer,&nbsp;Clémence Réda","doi":"10.1007/s10462-026-11656-w","DOIUrl":"10.1007/s10462-026-11656-w","url":null,"abstract":"<div><p>Over the past years, embedding learning on networks has shown tremendous results in link prediction tasks for complex systems, with a wide range of real-life applications. Learning a representation for each node in a knowledge graph allows us to capture topological and semantic information, which can be processed in downstream analyses later. In the link prediction task, high-dimensional network information is encoded into low-dimensional vectors, which are then fed to a predictor to infer new connections between nodes in the network. As the network complexity (that is, the numbers of connections and types of interactions) grows, embedding learning turns out increasingly challenging. This review covers published models on embedding learning on multiplex networks for link prediction. First, we propose refined taxonomies to classify and compare 38 models (in 36 papers) across five methodological categories and three embedding classes. Second, we review and address the problem of reproducible and fair evaluation of embedding learning on multiplex networks for the link prediction task. Finally, we highlight challenges specific to performance evaluation on directed multiplex networks, and address them by introducing a novel and fair testing procedure. This review constitutes a crucial step towards the development of more performant and tractable embedding learning approaches for multiplex networks and their fair evaluation for the link prediction task. We also suggest guidelines on the evaluation of models, and provide an informed perspective on the challenges and tools currently available to address downstream analyses applied to multiplex networks.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"59 9","pages":""},"PeriodicalIF":18.8,"publicationDate":"2026-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-026-11656-w.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148750687","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}