IoT data demands are growing, with Distributed Ledger Technologies (DLTs) offering secure data management, provided they can meet scaling and efficiency requirements that are more restrictive than in conventional application environments. This paper comprehensively surveys 27 DLTs of varying paradigms and implementation methods, proposes a scoring method for determining DLT-IoT integration suitability, and then applies that method to the surveyed DLTs. Six DLTs were shortlisted as the most promising, which were then subjected to in-depth analysis around three IoT use cases: health-IoT, e-commerce and automotive manufacturing. We discuss the viability of lightweight DLTs and identify crucial future research directions.
{"title":"Integration of IoT and Distributed Ledger Technologies: A Survey, Challenges, and Future Directions","authors":"Jusak Jusak, Steve Kerrison","doi":"10.1145/3789255","DOIUrl":"https://doi.org/10.1145/3789255","url":null,"abstract":"IoT data demands are growing, with Distributed Ledger Technologies (DLTs) offering secure data management, provided they can meet scaling and efficiency requirements that are more restrictive than in conventional application environments. This paper comprehensively surveys 27 DLTs of varying paradigms and implementation methods, proposes a scoring method for determining DLT-IoT integration suitability, and then applies that method to the surveyed DLTs. Six DLTs were shortlisted as the most promising, which were then subjected to in-depth analysis around three IoT use cases: health-IoT, e-commerce and automotive manufacturing. We discuss the viability of lightweight DLTs and identify crucial future research directions.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"20 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145986542","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}
Lianyu Hu, Mudi Jiang, Junjie Dong, Xinying Liu, Zengyou He
In recent years, much of the research on clustering algorithms has primarily focused on enhancing their accuracy and efficiency, frequently at the expense of interpretability. However, as these methods are increasingly being applied in high-stakes domains such as healthcare, finance, and autonomous systems, the need of transparent and interpretable clustering outcomes has become a critical concern. This is not only necessary for gaining user trust but also for satisfying the growing ethical and regulatory demands in these fields. Ensuring that decisions derived from clustering algorithms can be clearly understood and justified is now a fundamental requirement. To address this need, this paper provides a comprehensive and structured review of the current state of explainable clustering algorithms, identifying key criteria to distinguish between various methods. These insights can effectively assist researchers in making informed decisions about the most suitable explainable clustering methods for specific application contexts, while also promoting the development and adoption of clustering algorithms that are both efficient and transparent. For convenient access and reference, an open repository organizes representative and emerging interpretable clustering methods under the taxonomy proposed in this survey, available at https://github.com/hulianyu/Awesome-Interpretable-Clustering
{"title":"Interpretable Clustering: A Survey","authors":"Lianyu Hu, Mudi Jiang, Junjie Dong, Xinying Liu, Zengyou He","doi":"10.1145/3789495","DOIUrl":"https://doi.org/10.1145/3789495","url":null,"abstract":"In recent years, much of the research on clustering algorithms has primarily focused on enhancing their accuracy and efficiency, frequently at the expense of interpretability. However, as these methods are increasingly being applied in high-stakes domains such as healthcare, finance, and autonomous systems, the need of transparent and interpretable clustering outcomes has become a critical concern. This is not only necessary for gaining user trust but also for satisfying the growing ethical and regulatory demands in these fields. Ensuring that decisions derived from clustering algorithms can be clearly understood and justified is now a fundamental requirement. To address this need, this paper provides a comprehensive and structured review of the current state of explainable clustering algorithms, identifying key criteria to distinguish between various methods. These insights can effectively assist researchers in making informed decisions about the most suitable explainable clustering methods for specific application contexts, while also promoting the development and adoption of clustering algorithms that are both efficient and transparent. For convenient access and reference, an open repository organizes representative and emerging interpretable clustering methods under the taxonomy proposed in this survey, available at https://github.com/hulianyu/Awesome-Interpretable-Clustering","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"55 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145986496","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}
Kunal Chaturvedi, Nicholas Yang, Donald Dansereau, Christopher Lovejoy, Ali Braytee, Miriam Welgampola, Mukesh Prasad
Benign Positional Vertigo (BPV) is a common and correctable cause of dizziness worldwide, accompanied by unique nystagmus characteristics that can be recognized by trained healthcare workers. Nystagmus is an involuntary eye movement, consisting of an initial slow phase often followed by a subsequent quick phase, and is a key indicator of vestibular disorders including BPV. This review focuses on the application of machine learning Models for BPV diagnosis through the classification of nystagmus patterns. We examine the advancements in machine learning and deep learning techniques for nystagmus detection, highlighting the transition from traditional methods to more sophisticated approaches. We include a comprehensive analysis of recent studies, detailing the methodologies, datasets, and results of various models. We discuss the ongoing challenges and future directions in this domain, emphasizing the potential of these technologies to assist diagnosis of BPV by untrained clinicians and the promise of better patient outcomes. Through a systematic literature review process, this paper identifies gaps in current research and suggests areas for future exploration, aiming to support the application of artificial intelligence in the diagnosis of a common vertigo subtype.
{"title":"Diagnosis of Benign Positional Vertigo: A Systematic Review of Machine Learning and Deep Learning within Videonystagmography","authors":"Kunal Chaturvedi, Nicholas Yang, Donald Dansereau, Christopher Lovejoy, Ali Braytee, Miriam Welgampola, Mukesh Prasad","doi":"10.1145/3789494","DOIUrl":"https://doi.org/10.1145/3789494","url":null,"abstract":"Benign Positional Vertigo (BPV) is a common and correctable cause of dizziness worldwide, accompanied by unique nystagmus characteristics that can be recognized by trained healthcare workers. Nystagmus is an involuntary eye movement, consisting of an initial slow phase often followed by a subsequent quick phase, and is a key indicator of vestibular disorders including BPV. This review focuses on the application of machine learning Models for BPV diagnosis through the classification of nystagmus patterns. We examine the advancements in machine learning and deep learning techniques for nystagmus detection, highlighting the transition from traditional methods to more sophisticated approaches. We include a comprehensive analysis of recent studies, detailing the methodologies, datasets, and results of various models. We discuss the ongoing challenges and future directions in this domain, emphasizing the potential of these technologies to assist diagnosis of BPV by untrained clinicians and the promise of better patient outcomes. Through a systematic literature review process, this paper identifies gaps in current research and suggests areas for future exploration, aiming to support the application of artificial intelligence in the diagnosis of a common vertigo subtype.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"19 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145986544","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}
Li Huang, Sophie Ebersold, Alexander Kogtenkov, Bertrand Meyer, Yinling Liu
The technology of formal software verification has made spectacular advances, but how much does it actually benefit the development of practical software? Considerable disagreement remains about the practicality of building systems with mechanically-checked proofs of correctness. Is this prospect confined to a few expensive, life-critical projects, or can the idea be applied to a wide segment of the software industry? To help answer this question, the present survey examines a range of projects, in various application areas, that have produced formally verified systems and deployed them for actual use. It considers the technologies used, the form of verification applied, the results obtained, and the lessons that the software industry should draw regarding its ability to benefit from formal verification techniques and tools. Note: this version is the extended article, covering all the systems identified as relevant. A shorter version, covering only a selection, is also available.
{"title":"Lessons from Formally Verified Deployed Software Systems","authors":"Li Huang, Sophie Ebersold, Alexander Kogtenkov, Bertrand Meyer, Yinling Liu","doi":"10.1145/3785652","DOIUrl":"https://doi.org/10.1145/3785652","url":null,"abstract":"The technology of formal software verification has made spectacular advances, but how much does it actually benefit the development of practical software? Considerable disagreement remains about the practicality of building systems with mechanically-checked proofs of correctness. Is this prospect confined to a few expensive, life-critical projects, or can the idea be applied to a wide segment of the software industry? To help answer this question, the present survey examines a range of projects, in various application areas, that have produced formally verified systems and deployed them for actual use. It considers the technologies used, the form of verification applied, the results obtained, and the lessons that the software industry should draw regarding its ability to benefit from formal verification techniques and tools. Note: this version is the extended article, covering all the systems identified as relevant. A shorter version, covering only a selection, is also available.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"56 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145986543","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}
The swift evolution of Large Language Models (LLMs) like the GPT family, LLaMA, ChatGLM, and Qwen represents significant progress in artificial intelligence research. Despite their remarkable capabilities in generating content, these models encounter substantial challenges when producing structured outputs and engaging in dynamic interactions, particularly when they need to retrieve external information in real time. To address these limitations, researchers have developed the ”Function Calling” paradigm. This approach enables language models to analyze user inquiries and engage with defined functions, thereby facilitating precise responses through connections to external sources, including databases, programming interfaces, and live data streams. This functionality has been successfully implemented across numerous sectors such as finance analytics, healthcare systems, and service operations. The implementation of function calling comprises three essential phases: preparation, execution, and processing. The preparation phase encompasses query analysis and function identification. During execution, the system evaluates whether a function is necessary, extracts relevant parameters, and oversees the operation. The processing phase concentrates on analyzing outcomes and crafting appropriate responses. Each phase presents unique difficulties, ranging from accurately selecting functions to managing complex parameter extraction and ensuring reliable execution. Researchers have established various evaluation frameworks and metrics to assess function calling performance, including success rates, computational efficiency, parameter extraction accuracy, and response quality indicators such as ROUGE-L evaluation scores. This survey systematically reviews the current landscape of function calling in LLMs, analyzing technical challenges, examining existing solutions, and discussing evaluation methodologies. We particularly focus on practical implementations and industrial applications, providing insights into both current achievements and future directions in this rapidly evolving field. For a comprehensive collection of related research papers and the Appendix file, please refer to our repository at GitHub.
{"title":"Function Calling in Large Language Models: Industrial Practices, Challenges, and Future Directions","authors":"Maolin Wang, Yingyi Zhang, Bowen Yu, Bingguang Hao, Cunyin Peng, Yicheng Chen, Wei Zhou, Jinjie Gu, Chenyi Zhuang, Ruocheng Guo, Wanyu Wang, Xiangyu Zhao","doi":"10.1145/3788284","DOIUrl":"https://doi.org/10.1145/3788284","url":null,"abstract":"The swift evolution of Large Language Models (LLMs) like the GPT family, LLaMA, ChatGLM, and Qwen represents significant progress in artificial intelligence research. Despite their remarkable capabilities in generating content, these models encounter substantial challenges when producing structured outputs and engaging in dynamic interactions, particularly when they need to retrieve external information in real time. To address these limitations, researchers have developed the ”Function Calling” paradigm. This approach enables language models to analyze user inquiries and engage with defined functions, thereby facilitating precise responses through connections to external sources, including databases, programming interfaces, and live data streams. This functionality has been successfully implemented across numerous sectors such as finance analytics, healthcare systems, and service operations. The implementation of function calling comprises three essential phases: preparation, execution, and processing. The preparation phase encompasses query analysis and function identification. During execution, the system evaluates whether a function is necessary, extracts relevant parameters, and oversees the operation. The processing phase concentrates on analyzing outcomes and crafting appropriate responses. Each phase presents unique difficulties, ranging from accurately selecting functions to managing complex parameter extraction and ensuring reliable execution. Researchers have established various evaluation frameworks and metrics to assess function calling performance, including success rates, computational efficiency, parameter extraction accuracy, and response quality indicators such as ROUGE-L evaluation scores. This survey systematically reviews the current landscape of function calling in LLMs, analyzing technical challenges, examining existing solutions, and discussing evaluation methodologies. We particularly focus on practical implementations and industrial applications, providing insights into both current achievements and future directions in this rapidly evolving field. For a comprehensive collection of related research papers and the Appendix file, please refer to our repository at GitHub.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"57 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145968629","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}
Thomas Baumer, Sascha Kern, Ludwig Fuchs, Günther Pernul
Identity and Access Management (IAM) challenges organizations, requiring carefully orchestrated processes, technologies, and authorizations. Despite its strategic relevance, we lack a consolidated scientific understanding of IAM metrics and their alignment with IAM goals, like security, compliance, and operational efficiency. This systematic review aims to identify and classify IAM metrics from the literature to support evidence-based IAM. It links collected metrics to IAM goals and audiences. The literature review followed the guidelines of Levy and Ellis. It includes publications from databases SpringerLink, AIS eLibrary, IEEE Explore, ScienceDirect, ACM Digital Library, and relevant cross-referenced publications. The search strategy used keyword combinations, like ”Identity and Access Management” and ”Metrics,” since 2000. We screened and included publications based on eligibility criteria for relevance, quality, and the explicit presentation of IAM metrics, resulting in sixty publications. The review identified 43 IAM metrics, categorized by seven perspectives derived from IAM goals and processes. Each metric was analyzed by its target, impact on IAM goals, and relevant audiences. The synthesis shows that the literature lacks unified terminology and frameworks for IAM metrics. Future research includes standardizing terminology, linking metrics and targets to maturity levels, and establishing IAM process metrics. The DEVISE project funded this work. It was not registered in PROSPERO.
身份和访问管理（Identity and Access Management， IAM）对组织提出了挑战，需要精心编排流程、技术和授权。尽管它具有战略相关性，但我们对IAM指标及其与IAM目标（如安全性、合规性和运营效率）的一致性缺乏统一的科学理解。本系统综述旨在从文献中识别和分类IAM指标，以支持基于证据的IAM。它将收集到的指标与IAM目标和受众联系起来。文献综述遵循Levy和Ellis的指导方针。它包括来自数据库SpringerLink、AIS Library、IEEE Explore、ScienceDirect、ACM Digital Library和相关交叉引用出版物的出版物。自2000年以来，搜索策略使用关键字组合，如“身份和访问管理”和“度量”。我们根据相关性、质量和IAM指标的明确呈现的资格标准筛选并纳入了出版物，共纳入了60篇出版物。该综述确定了43个IAM指标，并从IAM目标和流程的7个角度进行了分类。每个指标都根据其目标、对IAM目标的影响和相关受众进行了分析。综合表明，文献缺乏统一的术语和框架的IAM指标。未来的研究包括标准化术语，将指标和目标与成熟度级别联系起来，以及建立IAM流程指标。设计项目资助了这项工作。它没有在普洛斯彼罗登记。
{"title":"Identity and Access Management Metrics: A Systematic Review","authors":"Thomas Baumer, Sascha Kern, Ludwig Fuchs, Günther Pernul","doi":"10.1145/3788858","DOIUrl":"https://doi.org/10.1145/3788858","url":null,"abstract":"Identity and Access Management (IAM) challenges organizations, requiring carefully orchestrated processes, technologies, and authorizations. Despite its strategic relevance, we lack a consolidated scientific understanding of IAM metrics and their alignment with IAM goals, like security, compliance, and operational efficiency. This systematic review aims to identify and classify IAM metrics from the literature to support evidence-based IAM. It links collected metrics to IAM goals and audiences. The literature review followed the guidelines of Levy and Ellis. It includes publications from databases SpringerLink, AIS eLibrary, IEEE Explore, ScienceDirect, ACM Digital Library, and relevant cross-referenced publications. The search strategy used keyword combinations, like ”Identity and Access Management” and ”Metrics,” since 2000. We screened and included publications based on eligibility criteria for relevance, quality, and the explicit presentation of IAM metrics, resulting in sixty publications. The review identified 43 IAM metrics, categorized by seven perspectives derived from IAM goals and processes. Each metric was analyzed by its target, impact on IAM goals, and relevant audiences. The synthesis shows that the literature lacks unified terminology and frameworks for IAM metrics. Future research includes standardizing terminology, linking metrics and targets to maturity levels, and establishing IAM process metrics. The DEVISE project funded this work. It was not registered in PROSPERO.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"4 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145968627","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}
Ratnabali Pal, Samarjit Kar, Dilip K. Prasad, Arif Ahmed Sekh
This survey explores the advancements in Visual Question Answering (VQA) technology designed for visually impaired (VI) individuals. VQA systems integrate computer vision (CV) and natural language processing (NLP). It has the potential to improve VI people’s independence by translating visual information into understandable representations. Compared to other recent surveys, this survey not only focuses on the development of the VQA system but also demonstrates the challenges faced by VI people. This paper provides an in-depth review of the state-of-the-art VQA systems, datasets, and methodologies, focusing on their application to assist VI users. We analyze the unique challenges faced by this demographic, such as the quality of images captured and the complexity of questions asked. The survey also highlights the specific needs of VI users and how existing VQA solutions address these needs. We discuss the role of multimodal transformers, prompt-based learning, and generative approaches in improving VQA performance. We discuss how it aligns with the Sustainable Development Goals (SDGs). Our findings emphasize the importance of developing customized VQA systems that meet the diverse requirements of VI individuals, leading the way for future research and innovation in this field. This comprehensive review aims to provide valuable insights and guidance for researchers and developers working on VQA technologies for VI people.
{"title":"VisionAidQA: Advancing Visual Question Answering for the Visually Impaired","authors":"Ratnabali Pal, Samarjit Kar, Dilip K. Prasad, Arif Ahmed Sekh","doi":"10.1145/3788861","DOIUrl":"https://doi.org/10.1145/3788861","url":null,"abstract":"This survey explores the advancements in Visual Question Answering (VQA) technology designed for visually impaired (VI) individuals. VQA systems integrate computer vision (CV) and natural language processing (NLP). It has the potential to improve VI people’s independence by translating visual information into understandable representations. Compared to other recent surveys, this survey not only focuses on the development of the VQA system but also demonstrates the challenges faced by VI people. This paper provides an in-depth review of the state-of-the-art VQA systems, datasets, and methodologies, focusing on their application to assist VI users. We analyze the unique challenges faced by this demographic, such as the quality of images captured and the complexity of questions asked. The survey also highlights the specific needs of VI users and how existing VQA solutions address these needs. We discuss the role of multimodal transformers, prompt-based learning, and generative approaches in improving VQA performance. We discuss how it aligns with the Sustainable Development Goals (SDGs). Our findings emphasize the importance of developing customized VQA systems that meet the diverse requirements of VI individuals, leading the way for future research and innovation in this field. This comprehensive review aims to provide valuable insights and guidance for researchers and developers working on VQA technologies for VI people.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"30 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145968628","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}
Chen He, Niklas Elmqvist, Andrea Bellucci, Mahdi Munshi, Giulio Jacucci
Insight generation has long been a primary goal of information visualization. Yet, there remains a lack of a comprehensive review on the research advancements of Visualization Insight. This systematic literature review analyzes nearly two decades of theoretical, empirical, and technological research development of visualization insight. By examining 340 articles from 17 leading journals and conferences in visualization, human-computer interaction, and databases and conducting an in-depth review of 92 papers, we propose a framework illustrating the iterative and collaborative process of generating, exploring, and communicating visualization insights. We then use the framework to analyze and compare the reviewed literature, revealing that 1) Research on visualization insight has evolved from defining insights to formalizing the insight generation process and communicating visualization insights through reports and data stories; 2) Issues like cognitive biases and chart errors hinder the effectiveness of visual discovery; 3) Automation in data fact mining and story generation eases data storytelling, but often lacks the incorporation of domain information, limiting the stories’ impact. Moving forward, we outline five future research directions, aiming to improve insight quality and expand the methods of generating and communicating insights.
{"title":"Systemization of Knowledge (SoK): Visualization Insight -- Two Decades of Research, Practice, and Future Directions","authors":"Chen He, Niklas Elmqvist, Andrea Bellucci, Mahdi Munshi, Giulio Jacucci","doi":"10.1145/3788869","DOIUrl":"https://doi.org/10.1145/3788869","url":null,"abstract":"Insight generation has long been a primary goal of information visualization. Yet, there remains a lack of a comprehensive review on the research advancements of Visualization Insight. This systematic literature review analyzes nearly two decades of theoretical, empirical, and technological research development of visualization insight. By examining 340 articles from 17 leading journals and conferences in visualization, human-computer interaction, and databases and conducting an in-depth review of 92 papers, we propose a framework illustrating the iterative and collaborative process of generating, exploring, and communicating visualization insights. We then use the framework to analyze and compare the reviewed literature, revealing that 1) Research on visualization insight has evolved from defining insights to formalizing the insight generation process and communicating visualization insights through reports and data stories; 2) Issues like cognitive biases and chart errors hinder the effectiveness of visual discovery; 3) Automation in data fact mining and story generation eases data storytelling, but often lacks the incorporation of domain information, limiting the stories’ impact. Moving forward, we outline five future research directions, aiming to improve insight quality and expand the methods of generating and communicating insights.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"53 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145968631","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}
Algorithm design is crucial for effective problem-solving across various domains. The advent of Large Language Models (LLMs) has notably enhanced the automation and innovation within this field, offering new perspectives and promising solutions. In just a few years, this integration has yielded remarkable progress in areas ranging from combinatorial optimization to scientific discovery. Despite this rapid expansion, a holistic understanding of the field is hindered by the lack of a systematic review, as existing surveys either remain limited to narrow sub-fields or with different objectives. This paper seeks to provide a systematic review of algorithm design with LLMs. We introduce a taxonomy that categorises the roles of LLMs as optimizers, predictors, extractors and designers, analyzing the progress, advantages, and limitations within each category. We further synthesize literature across the three phases of the algorithm design pipeline and across diverse algorithmic applications that define the current landscape. Finally, we outline key open challenges and opportunities to guide future research.
{"title":"A Systematic Survey on Large Language Models for Algorithm Design","authors":"Fei Liu, Yiming Yao, Ping Guo, Zhiyuan Yang, Xi Lin, Zhe Zhao, Xialiang Tong, Kun Mao, Zhichao Lu, Zhenkun Wang, Mingxuan Yuan, Qingfu Zhang","doi":"10.1145/3787585","DOIUrl":"https://doi.org/10.1145/3787585","url":null,"abstract":"Algorithm design is crucial for effective problem-solving across various domains. The advent of Large Language Models (LLMs) has notably enhanced the automation and innovation within this field, offering new perspectives and promising solutions. In just a few years, this integration has yielded remarkable progress in areas ranging from combinatorial optimization to scientific discovery. Despite this rapid expansion, a holistic understanding of the field is hindered by the lack of a systematic review, as existing surveys either remain limited to narrow sub-fields or with different objectives. This paper seeks to provide a systematic review of algorithm design with LLMs. We introduce a taxonomy that categorises the roles of LLMs as optimizers, predictors, extractors and designers, analyzing the progress, advantages, and limitations within each category. We further synthesize literature across the three phases of the algorithm design pipeline and across diverse algorithmic applications that define the current landscape. Finally, we outline key open challenges and opportunities to guide future research.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"29 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145955296","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}
Anomaly detection has been a cornerstone of research across diverse disciplines, remaining a critical and evolving field of study for several decades. While several approaches, including deep learning (DL), have been explored to design general solutions for anomaly detection, there is a lack of structured survey articles on the collective (or group) anomaly detection problem (CAD). This article fills this gap and presents the first comprehensive review dedicated exclusively to DL-based methods for CAD. A two-level taxonomy is proposed to categorize CAD methods based on their underlying algorithms into generative, discriminative, or hybrid, and the methods are further classified according to their DL architecture. The literature review we conducted on CAD reveals that the existing approaches utilizing deep learning address a wide range of applications, including cybersecurity, IoT, and key performance indicators (KPIs). Different application domains of CAD and benchmark datasets are discussed. Moreover, the commonly used datasets for CAD are described and discussed from different application scenarios. Finally, the limitations and drawbacks of the different trends used to solve the problem of CAD are outlined, and the challenges and future research directions are discussed.
{"title":"Deep Learning for Collective Anomaly Detection","authors":"Dalila Khettaf, Djamel Djenouri, Zeinab Rezaeifar, Youcef Djenouri","doi":"10.1145/3788280","DOIUrl":"https://doi.org/10.1145/3788280","url":null,"abstract":"Anomaly detection has been a cornerstone of research across diverse disciplines, remaining a critical and evolving field of study for several decades. While several approaches, including deep learning (DL), have been explored to design general solutions for anomaly detection, there is a lack of structured survey articles on the collective (or group) anomaly detection problem (CAD). This article fills this gap and presents the first comprehensive review dedicated exclusively to DL-based methods for CAD. A two-level taxonomy is proposed to categorize CAD methods based on their underlying algorithms into generative, discriminative, or hybrid, and the methods are further classified according to their DL architecture. The literature review we conducted on CAD reveals that the existing approaches utilizing deep learning address a wide range of applications, including cybersecurity, IoT, and key performance indicators (KPIs). Different application domains of CAD and benchmark datasets are discussed. Moreover, the commonly used datasets for CAD are described and discussed from different application scenarios. Finally, the limitations and drawbacks of the different trends used to solve the problem of CAD are outlined, and the challenges and future research directions are discussed.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"2 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145947218","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：