This survey investigates the multifaceted nature of selective forgetting in machine learning, drawing insights from neuroscientific research that posits forgetting as an adaptive function rather than a defect, enhancing the learning process and preventing overfitting. This survey focuses on the benefits of selective forgetting and its applications across various machine learning sub-fields that can help improve model performance and enhance data privacy. Moreover, the paper discusses current challenges, future directions, and ethical considerations regarding the integration of selective forgetting mechanisms into machine learning models. We present a comprehensive taxonomy that bridges diverse selective forgetting-related research in machine learning, systematically categorising approaches along key dimensions. Our work synthesises theories of forgetting from different knowledge areas to establish theoretical foundations for forgetting mechanisms in machine learning, providing a unified framework for understanding selective forgetting processes.
{"title":"Selective Forgetting in Machine Learning and Beyond: A Survey","authors":"Alyssa Sha, Bernardo Nunes, Armin Haller","doi":"10.1145/3796542","DOIUrl":"https://doi.org/10.1145/3796542","url":null,"abstract":"This survey investigates the multifaceted nature of selective forgetting in machine learning, drawing insights from neuroscientific research that posits forgetting as an adaptive function rather than a defect, enhancing the learning process and preventing overfitting. This survey focuses on the benefits of selective forgetting and its applications across various machine learning sub-fields that can help improve model performance and enhance data privacy. Moreover, the paper discusses current challenges, future directions, and ethical considerations regarding the integration of selective forgetting mechanisms into machine learning models. We present a comprehensive taxonomy that bridges diverse selective forgetting-related research in machine learning, systematically categorising approaches along key dimensions. Our work synthesises theories of forgetting from different knowledge areas to establish theoretical foundations for forgetting mechanisms in machine learning, providing a unified framework for understanding selective forgetting processes.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"186 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146153680","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Buhua Liu, Shitong Shao, Bao Li, Lichen Bai, Zhiqiang Xu, Haoyi Xiong, James T. Kwok, Sumi Helal, Zeke Xie
Diffusion models have emerged as the leading paradigm in generative modeling, excelling in various applications. Despite their success, these models often misalign with human intentions and generate results with undesired properties or even harmful content. Inspired by the success and popularity of alignment in tuning large language models, recent studies have investigated aligning diffusion models with human expectations and preferences. This work mainly reviews alignment of diffusion models, covering advancements in fundamentals of alignment, alignment techniques of diffusion models, preference benchmarks, and evaluation for diffusion models. Moreover, we discuss key perspectives on current challenges and promising future directions on solving the remaining challenges in alignment of diffusion models. To the best of our knowledge, our work is the first comprehensive review paper for researchers and engineers to comprehend, practice, and research alignment of diffusion models.
{"title":"Alignment of Diffusion Models: Fundamentals, Challenges, and Future","authors":"Buhua Liu, Shitong Shao, Bao Li, Lichen Bai, Zhiqiang Xu, Haoyi Xiong, James T. Kwok, Sumi Helal, Zeke Xie","doi":"10.1145/3796982","DOIUrl":"https://doi.org/10.1145/3796982","url":null,"abstract":"Diffusion models have emerged as the leading paradigm in generative modeling, excelling in various applications. Despite their success, these models often misalign with human intentions and generate results with undesired properties or even harmful content. Inspired by the success and popularity of alignment in tuning large language models, recent studies have investigated aligning diffusion models with human expectations and preferences. This work mainly reviews alignment of diffusion models, covering advancements in fundamentals of alignment, alignment techniques of diffusion models, preference benchmarks, and evaluation for diffusion models. Moreover, we discuss key perspectives on current challenges and promising future directions on solving the remaining challenges in alignment of diffusion models. To the best of our knowledge, our work is the first comprehensive review paper for researchers and engineers to comprehend, practice, and research alignment of diffusion models.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"26 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146153662","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Although Deep Neural Networks (DNNs) have been widely applied in various real-world scenarios, they remain vulnerable to adversarial examples. Adversarial attacks in computer vision can be categorized into digital attacks and physical attacks based on their different forms. Compared to digital attacks, which generate perturbations in digital pixels, physical attacks are more practical in real-world settings. Due to the serious security risks posed by physically adversarial examples, many studies have been conducted to evaluate the physically adversarial robustness of DNNs in recent years. In this paper, we provide a comprehensive survey of current physically adversarial attacks and defenses in computer vision. We establish a taxonomy by organizing physical attacks according to attack tasks, attack forms, and attack methods. This approach offers readers a systematic understanding of the topic from multiple perspectives. For physical defenses, we categorize them into pre-processing, in-processing, and post-processing for DNN models to ensure comprehensive coverage of adversarial defenses. Based on this survey, we discuss the challenges facing this research field and provide an outlook on future directions.
{"title":"Visual Adversarial Attacks and Defenses in the Physical World: A Survey","authors":"Xingxing Wei, Bangzheng Pu, Shiji Zhao, Jiefan Lu, Baoyuan Wu","doi":"10.1145/3793659","DOIUrl":"https://doi.org/10.1145/3793659","url":null,"abstract":"Although Deep Neural Networks (DNNs) have been widely applied in various real-world scenarios, they remain vulnerable to adversarial examples. Adversarial attacks in computer vision can be categorized into digital attacks and physical attacks based on their different forms. Compared to digital attacks, which generate perturbations in digital pixels, physical attacks are more practical in real-world settings. Due to the serious security risks posed by physically adversarial examples, many studies have been conducted to evaluate the physically adversarial robustness of DNNs in recent years. In this paper, we provide a comprehensive survey of current physically adversarial attacks and defenses in computer vision. We establish a taxonomy by organizing physical attacks according to attack tasks, attack forms, and attack methods. This approach offers readers a systematic understanding of the topic from multiple perspectives. For physical defenses, we categorize them into pre-processing, in-processing, and post-processing for DNN models to ensure comprehensive coverage of adversarial defenses. Based on this survey, we discuss the challenges facing this research field and provide an outlook on future directions.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"299 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146153645","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jiacheng Liu, Peng Tang, Wenfeng Wang, Yuhang Ren, Xiaofeng Hou, Pheng Ann Heng, Minyi Guo, Chao Li
The emergence of large-scale Mixture of Experts (MoE) models represents a significant advancement in artificial intelligence, offering larger model capacity and computational efficiency through conditional computation. However, deploying and running inference on these models presents significant challenges in computational resources, latency, and energy efficiency. This comprehensive survey analyzes optimization techniques for MoE models across the entire system stack. We first establish a taxonomical framework that categorizes optimization approaches into model-level, system-level, and hardware-level optimizations. At the model level, we examine architectural innovations including efficient expert design, attention mechanisms, various compression techniques such as pruning, quantization, and knowledge distillation, as well as algorithm improvement including dynamic routing strategies and expert merging methods. At the system level, we investigate distributed computing approaches, load balancing mechanisms, and efficient scheduling algorithms that enable scalable deployment. Furthermore, we delve into hardware-specific optimizations and co-design strategies that maximize throughput and energy efficiency. This survey provides both a structured overview of existing solutions and identifies key challenges and promising research directions in MoE inference optimization. To facilitate ongoing updates and the sharing of cutting-edge advances in MoE inference optimization research, we have established a repository accessible at https://github.com/MoE-Inf/awesome-moe-inference/.
{"title":"A Survey on Inference Optimization Techniques for Mixture of Experts Models","authors":"Jiacheng Liu, Peng Tang, Wenfeng Wang, Yuhang Ren, Xiaofeng Hou, Pheng Ann Heng, Minyi Guo, Chao Li","doi":"10.1145/3794845","DOIUrl":"https://doi.org/10.1145/3794845","url":null,"abstract":"The emergence of large-scale Mixture of Experts (MoE) models represents a significant advancement in artificial intelligence, offering larger model capacity and computational efficiency through conditional computation. However, deploying and running inference on these models presents significant challenges in computational resources, latency, and energy efficiency. This comprehensive survey analyzes optimization techniques for MoE models across the entire system stack. We first establish a taxonomical framework that categorizes optimization approaches into model-level, system-level, and hardware-level optimizations. At the model level, we examine architectural innovations including efficient expert design, attention mechanisms, various compression techniques such as pruning, quantization, and knowledge distillation, as well as algorithm improvement including dynamic routing strategies and expert merging methods. At the system level, we investigate distributed computing approaches, load balancing mechanisms, and efficient scheduling algorithms that enable scalable deployment. Furthermore, we delve into hardware-specific optimizations and co-design strategies that maximize throughput and energy efficiency. This survey provides both a structured overview of existing solutions and identifies key challenges and promising research directions in MoE inference optimization. To facilitate ongoing updates and the sharing of cutting-edge advances in MoE inference optimization research, we have established a repository accessible at https://github.com/MoE-Inf/awesome-moe-inference/.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"4 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146153664","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Zhenzhong Wang, Haowei Hua, Wanyu Lin, Ming Yang, Kay Chen Tan
Crystalline materials, with symmetrical and periodic structures, exhibit a wide spectrum of properties and have been widely used in numerous applications across electronics, energy, and beyond. For crystalline materials discovery, traditional experimental and computational approaches are time-consuming and expensive. In these years, thanks to the explosive amount of crystalline materials data, great interest has been given to data-driven materials discovery. Particularly, recent advancements have exploited the expressive representation ability of deep learning to model the highly complex atomic systems within crystalline materials, opening up new avenues for efficient and accurate materials discovery. These works main focus on four types of tasks, including physicochemical property prediction, generative design of crystalline materials, aiding characterization, and accelerating theoretical computations. Despite the remarkable progress, there is still a lack of systematic investigation to summarize their distinctions and limitations. To fill this gap, we systematically investigated the progress of crystalline materials discovery using artificial intelligence made in recent years. We first introduce several data representations of the crystalline materials. Based on the representations, we summarize various fundamental deep learning models and their tailored usages in various material discovery tasks. Finally, we highlight the remaining challenges and propose future directions.
{"title":"Crystalline Material Discovery in the Era of Artificial Intelligence","authors":"Zhenzhong Wang, Haowei Hua, Wanyu Lin, Ming Yang, Kay Chen Tan","doi":"10.1145/3794853","DOIUrl":"https://doi.org/10.1145/3794853","url":null,"abstract":"Crystalline materials, with symmetrical and periodic structures, exhibit a wide spectrum of properties and have been widely used in numerous applications across electronics, energy, and beyond. For crystalline materials discovery, traditional experimental and computational approaches are time-consuming and expensive. In these years, thanks to the explosive amount of crystalline materials data, great interest has been given to data-driven materials discovery. Particularly, recent advancements have exploited the expressive representation ability of deep learning to model the highly complex atomic systems within crystalline materials, opening up new avenues for efficient and accurate materials discovery. These works main focus on four types of tasks, including physicochemical property prediction, generative design of crystalline materials, aiding characterization, and accelerating theoretical computations. Despite the remarkable progress, there is still a lack of systematic investigation to summarize their distinctions and limitations. To fill this gap, we systematically investigated the progress of crystalline materials discovery using artificial intelligence made in recent years. We first introduce several data representations of the crystalline materials. Based on the representations, we summarize various fundamental deep learning models and their tailored usages in various material discovery tasks. Finally, we highlight the remaining challenges and propose future directions.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"71 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146153681","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this study, we evaluate potential demographic bias in state-of-the-art deepfake image detection models across three key attributes: age, ethnicity, and gender. Unlike prior works that retrain detectors or analyse forensic manipulations, we systematically assess multiple pretrained checkpoints of leading deepfake detectors, each trained on different datasets, to ensure an unbiased evaluation framework. Our experiments employ synthetic images generated by recent diffusion and autoregressive models, alongside real images from balanced datasets, to measure subgroup-specific detection performance. Results reveal no systematic bias across demographic categories—variations in accuracy and precision remain within small statistical margins across all detectors and checkpoints. We further provide a taxonomy of image generative models, highlighting their evolution from pixel-space to latent-space diffusion architectures, to contextualize the diversity of synthetic data used in our evaluation. Overall, our findings suggest that modern deepfake image detectors, when tested in a cross-demographic setting using pretrained checkpoints, exhibit robust and fair performance across age, ethnicity, and gender.
{"title":"Bias-Free? An Empirical Study on Ethnicity, Gender, and Age Fairness in Deepfake Detection","authors":"Aditi Panda, Tanusree Ghosh, Tushar Choudhary, Ruchira Naskar","doi":"10.1145/3796544","DOIUrl":"https://doi.org/10.1145/3796544","url":null,"abstract":"In this study, we evaluate potential demographic bias in state-of-the-art deepfake image detection models across three key attributes: age, ethnicity, and gender. Unlike prior works that retrain detectors or analyse forensic manipulations, we systematically assess multiple pretrained checkpoints of leading deepfake detectors, each trained on different datasets, to ensure an unbiased evaluation framework. Our experiments employ synthetic images generated by recent diffusion and autoregressive models, alongside real images from balanced datasets, to measure subgroup-specific detection performance. Results reveal no systematic bias across demographic categories—variations in accuracy and precision remain within small statistical margins across all detectors and checkpoints. We further provide a taxonomy of image generative models, highlighting their evolution from pixel-space to latent-space diffusion architectures, to contextualize the diversity of synthetic data used in our evaluation. Overall, our findings suggest that modern deepfake image detectors, when tested in a cross-demographic setting using pretrained checkpoints, exhibit robust and fair performance across age, ethnicity, and gender.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"40 4 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146153682","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Domain ontologies are essential for representing and reasoning about knowledge, yet addressing information uncertainty within them remains challenging. This review surveys approaches to modelling information uncertainty in domain ontologies from 2010 to 2024. It categorizes modelling formalisms, identifies information uncertainty types, and analyzes how information uncertainty is integrated into ontology components. It reviews reasoning techniques and emerging methods, including Machine Learning and Natural Language Processing. The review examines languages, tools, and evaluation strategies. The purpose is to map the landscape of information uncertainty modelling in domain ontologies, highlight research gaps and trends, and provide structured guidance for selecting suitable approaches.
{"title":"A Comprehensive Review of Information Uncertainty Modelling in Domain Ontologies","authors":"Deemah Alomair, Ridha Khedri, Wendy MacCaull","doi":"10.1145/3794841","DOIUrl":"https://doi.org/10.1145/3794841","url":null,"abstract":"Domain ontologies are essential for representing and reasoning about knowledge, yet addressing information uncertainty within them remains challenging. This review surveys approaches to modelling information uncertainty in domain ontologies from 2010 to 2024. It categorizes modelling formalisms, identifies information uncertainty types, and analyzes how information uncertainty is integrated into ontology components. It reviews reasoning techniques and emerging methods, including Machine Learning and Natural Language Processing. The review examines languages, tools, and evaluation strategies. The purpose is to map the landscape of information uncertainty modelling in domain ontologies, highlight research gaps and trends, and provide structured guidance for selecting suitable approaches.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"41 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146135118","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Minxing Zhang, Yi Yang, Roy Xie, Bhuwan Dhingra, Shuyan Zhou, Jian Pei
Large Language Model (LLM)-based agents have recently emerged as a new paradigm that extends the capabilities of LLMs beyond text generation to dynamic interaction with external environments. A critical challenge lies in ensuring their generalizability – the ability to maintain consistently high performance across varied instructions, tasks, environments, and domains, especially those different from the agent’s fine-tuning data. Despite growing interest, the concept of generalizability in LLM-based agents remains underdefined, and systematic approaches to measure and improve it are lacking. We provide the first comprehensive review of generalizability in LLM-based agents. We begin by clarifying the definition and boundaries of agent generalizability. We then review existing benchmarks. Next, we categorize strategies for improving generalizability into three groups: methods targeting the backbone LLM, targeting agent components, and targeting their interactions. Furthermore, we introduce the distinction between generalizable frameworks and generalizable agents and outline how generalizable frameworks can be translated into agent-level generalizability. Finally, we identify future directions, including the development of standardized evaluation frameworks, variance- and cost-based metrics, and hybrid approaches that integrate methodological innovations with agent architecture-level designs. We aim to establish a foundation for principled research on building LLM-based agents that generalize reliably across diverse real-world applications.
{"title":"Generalizability of Large Language Model-Based Agents: A Comprehensive Survey","authors":"Minxing Zhang, Yi Yang, Roy Xie, Bhuwan Dhingra, Shuyan Zhou, Jian Pei","doi":"10.1145/3794858","DOIUrl":"https://doi.org/10.1145/3794858","url":null,"abstract":"Large Language Model (LLM)-based agents have recently emerged as a new paradigm that extends the capabilities of LLMs beyond text generation to dynamic interaction with external environments. A critical challenge lies in ensuring their <jats:italic toggle=\"yes\">generalizability</jats:italic> – the ability to maintain consistently high performance across varied instructions, tasks, environments, and domains, especially those different from the agent’s fine-tuning data. Despite growing interest, the concept of generalizability in LLM-based agents remains underdefined, and systematic approaches to measure and improve it are lacking. We provide the first comprehensive review of generalizability in LLM-based agents. We begin by clarifying the definition and boundaries of agent generalizability. We then review existing benchmarks. Next, we categorize strategies for improving generalizability into three groups: methods targeting the backbone LLM, targeting agent components, and targeting their interactions. Furthermore, we introduce the distinction between <jats:italic toggle=\"yes\">generalizable frameworks</jats:italic> and <jats:italic toggle=\"yes\">generalizable agents</jats:italic> and outline how generalizable frameworks can be translated into agent-level generalizability. Finally, we identify future directions, including the development of standardized evaluation frameworks, variance- and cost-based metrics, and hybrid approaches that integrate methodological innovations with agent architecture-level designs. We aim to establish a foundation for principled research on building LLM-based agents that generalize reliably across diverse real-world applications.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"244 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146138700","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This meta-analysis examines music emotion recognition (MER) models published between 2014 and 2024, focusing on predictions of valence, arousal, and categorical emotions. A total of 553 studies were identified, of which 96 full-text articles were assessed, resulting in a final review of 34 studies. These studies reported 290 models, including 86 for emotion classification and 204 for regression. Using the best-performing model from each study, we found that valence and arousal were predicted with reasonable accuracy (r = 0.67 and r = 0.81, respectively), while classification models achieved an accuracy of 0.87 as measured with Matthews correlation coefficient. Across modelling approaches, linear and tree-based methods generally outperformed neural networks in regression tasks, whereas neural networks and support vector machines (SVMs) showed highest performance in classification tasks. We highlight key recommendations for future MER research, emphasizing the need for greater transparency, feature validation, and standardized reporting to improve comparability across studies.
{"title":"A Meta-Analysis of Music Emotion Recognition Studies","authors":"Tuomas Eerola, Cameron Anderson","doi":"10.1145/3796518","DOIUrl":"https://doi.org/10.1145/3796518","url":null,"abstract":"This meta-analysis examines music emotion recognition (MER) models published between 2014 and 2024, focusing on predictions of valence, arousal, and categorical emotions. A total of 553 studies were identified, of which 96 full-text articles were assessed, resulting in a final review of 34 studies. These studies reported 290 models, including 86 for emotion classification and 204 for regression. Using the best-performing model from each study, we found that valence and arousal were predicted with reasonable accuracy (r = 0.67 and r = 0.81, respectively), while classification models achieved an accuracy of 0.87 as measured with Matthews correlation coefficient. Across modelling approaches, linear and tree-based methods generally outperformed neural networks in regression tasks, whereas neural networks and support vector machines (SVMs) showed highest performance in classification tasks. We highlight key recommendations for future MER research, emphasizing the need for greater transparency, feature validation, and standardized reporting to improve comparability across studies.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"72 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146138871","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Philipp Scharpf, Corinna Breitinger, Andreas Spitz, Norman Meuschke, André Greiner-Petter, Moritz Schubotz, Bela Gipp
This article provides a comprehensive systematic review of the literature on entity linking using Wikidata as the grounding knowledge base. Our review extends the scope of previous studies from two to eight dimensions of entity linking, which we classify into the following categories: definitions, tasks, types, domains, approaches, datasets, applications, and challenges. We find that datasets primarily address question-answering and news domains but underutilize Wikidata’s capabilities for hyper-relations, multilingualism, and time dependence. The research gaps we identify include the need for more robust datasets, hybrid methods combining rule-based and learning-based approaches, and improved handling of ambiguity, sparse entity types, data noise, and knowledge graph evolution.
{"title":"Entity Linking with Wikidata: A Systematic Literature Review","authors":"Philipp Scharpf, Corinna Breitinger, Andreas Spitz, Norman Meuschke, André Greiner-Petter, Moritz Schubotz, Bela Gipp","doi":"10.1145/3795134","DOIUrl":"https://doi.org/10.1145/3795134","url":null,"abstract":"This article provides a comprehensive systematic review of the literature on entity linking using Wikidata as the grounding knowledge base. Our review extends the scope of previous studies from two to eight dimensions of entity linking, which we classify into the following categories: definitions, tasks, types, domains, approaches, datasets, applications, and challenges. We find that datasets primarily address question-answering and news domains but underutilize Wikidata’s capabilities for hyper-relations, multilingualism, and time dependence. The research gaps we identify include the need for more robust datasets, hybrid methods combining rule-based and learning-based approaches, and improved handling of ambiguity, sparse entity types, data noise, and knowledge graph evolution.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"9 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146095841","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：