Network middleboxes are important components in modern networking systems, impacting approximately 40% of network paths according to recent studies [1]. This survey paper delves into their endemic presence, enriches the original 2002 RFC with over two decades of findings, and emphasizes the significance of their impact in terms of security and performance. Furthermore, it categorizes network middleboxes based on their functions, objectives, and alterations. In today’s world, network middleboxes emerge as a dual-edged sword. While important for network operations, they also pose security risks. We present the various challenges they introduce, including their contribution to Internet ossification, their potential for censorship, monitoring, and traffic differentiation. Substantial effort remains to make their presence more visible to end users. This paper explores potential solutions, ranging from prevention and detection to curative measures. Ultimately, we aim to establish this survey as a foundational resource for addressing challenges revolving around the notion of network middleboxes, thereby fostering further research and innovation in this area.
{"title":"ENDEMIC: End-to-End Network Disruptions - Examining Middleboxes, Issues, and Countermeasures - A Survey","authors":"Ilies Benhabbour, Marc Dacier","doi":"10.1145/3716372","DOIUrl":"https://doi.org/10.1145/3716372","url":null,"abstract":"Network middleboxes are important components in modern networking systems, impacting approximately 40% of network paths according to recent studies [1]. This survey paper delves into their endemic presence, enriches the original 2002 RFC with over two decades of findings, and emphasizes the significance of their impact in terms of security and performance. Furthermore, it categorizes network middleboxes based on their functions, objectives, and alterations. In today’s world, network middleboxes emerge as a dual-edged sword. While important for network operations, they also pose security risks. We present the various challenges they introduce, including their contribution to Internet ossification, their potential for censorship, monitoring, and traffic differentiation. Substantial effort remains to make their presence more visible to end users. This paper explores potential solutions, ranging from prevention and detection to curative measures. Ultimately, we aim to establish this survey as a foundational resource for addressing challenges revolving around the notion of network middleboxes, thereby fostering further research and innovation in this area.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"136 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2025-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143191817","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}
Ahmad Jan Mian, Muhammad Adil, Bouziane Brik, Saad Harous, Sohail Abbas
Intelligent Transportation Systems (ITS) generate massive amounts of Big Data through both sensory and non-sensory platforms. The data support batch processing as well as stream processing, which are essential for reliable operations on the roads and connected vehicles in ITS. Despite the immense potential of Big Data intelligence in ITS, autonomous vehicles are largely confined to testing and trial phases. The research community is working tirelessly to improve the reliability of ITS by designing new protocols, standards and connectivity paradigms. In the recent past, several surveys have been conducted that focus on Big Data Intelligence for ITS, yet none of them have comprehensively addressed the fundamental challenges hindering the widespread adoption of autonomous vehicles on the roads. Our survey aims to help readers better understand the technological advancements by delving deep into Big Data architecture, focusing on data acquisition, data storage and data visualization. We reviewed sensory and non-sensory platforms for data acquisition, data storage repositories for archival and retrieval of large datasets, and data visualization for presenting the processed data in an interactive and comprehensible format. To this end, we discussed the current research progress by comprehensively covering the literature and highlighting challenges that urgently require the attention of research community. Based on the concluding remarks, we argued that these challenges hinder the widespread presence of autonomous vehicles on the roads. Understanding these challenges is important for a more informed discussion on the future of self-driven technology. Moreover, we acknowledge that these challenges not only affect individual layers but also impact the functionality of subsequent layers. Finally, we outline our future work that explores how resolving these challenges could enable the realization of innovations such as smart charging systems on the roads and data centers on wheels.
{"title":"Making Sense of Big Data in Intelligent Transportation Systems: Current Trends, Challenges and Future Directions","authors":"Ahmad Jan Mian, Muhammad Adil, Bouziane Brik, Saad Harous, Sohail Abbas","doi":"10.1145/3716371","DOIUrl":"https://doi.org/10.1145/3716371","url":null,"abstract":"Intelligent Transportation Systems (ITS) generate massive amounts of Big Data through both sensory and non-sensory platforms. The data support batch processing as well as stream processing, which are essential for reliable operations on the roads and connected vehicles in ITS. Despite the immense potential of Big Data intelligence in ITS, autonomous vehicles are largely confined to testing and trial phases. The research community is working tirelessly to improve the reliability of ITS by designing new protocols, standards and connectivity paradigms. In the recent past, several surveys have been conducted that focus on Big Data Intelligence for ITS, yet none of them have comprehensively addressed the fundamental challenges hindering the widespread adoption of autonomous vehicles on the roads. Our survey aims to help readers better understand the technological advancements by delving deep into Big Data architecture, focusing on data acquisition, data storage and data visualization. We reviewed sensory and non-sensory platforms for data acquisition, data storage repositories for archival and retrieval of large datasets, and data visualization for presenting the processed data in an interactive and comprehensible format. To this end, we discussed the current research progress by comprehensively covering the literature and highlighting challenges that urgently require the attention of research community. Based on the concluding remarks, we argued that these challenges hinder the widespread presence of autonomous vehicles on the roads. Understanding these challenges is important for a more informed discussion on the future of self-driven technology. Moreover, we acknowledge that these challenges not only affect individual layers but also impact the functionality of subsequent layers. Finally, we outline our future work that explores how resolving these challenges could enable the realization of innovations such as smart charging systems on the roads and data centers on wheels.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"62 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2025-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143191818","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}
Artificial Intelligence (AI) fosters enormous business opportunities that build and utilize private AI models. Implementing AI models at scale and ensuring cost-effective production of AI-based technologies through entirely in-house capabilities is a challenge. The success of the Infrastructure as a Service (IaaS) and Software as a Service (SaaS) Cloud Computing models can be leveraged to facilitate a cost-effective and scalable AI service paradigm, namely, ‘AI as a Service.’ We summarize current state-of-the-art solutions for AI-as-a-Service (AIaaS), and we discuss its prospects for growth and opportunities to advance the concept. To this end, we perform a thorough review of recent research on AI and various deployment strategies for emerging domains considering both technical as well as survey articles. Next, we identify various characteristics and capabilities that need to be met before an AIaaS model can be successfully designed and deployed. Based on this we present a general framework of an AIaaS architecture that integrates the required aaS characteristics with the capabilities of AI. We also compare various approaches for offering AIaaS to end users. Finally, we illustrate several real-world use cases for AIaaS models, followed by a discussion of some of the challenges that must be addressed to enable AIaaS adoption.
{"title":"Artificial Intelligence as a Service (AIaaS) for Cloud, Fog and the Edge: State-of-the-Art Practices","authors":"Naeem Syed, Adnan Anwar, Zubair Baig, Sherali Zeadally","doi":"10.1145/3712016","DOIUrl":"https://doi.org/10.1145/3712016","url":null,"abstract":"Artificial Intelligence (AI) fosters enormous business opportunities that build and utilize private AI models. Implementing AI models at scale and ensuring cost-effective production of AI-based technologies through entirely in-house capabilities is a challenge. The success of the Infrastructure as a Service (IaaS) and Software as a Service (SaaS) Cloud Computing models can be leveraged to facilitate a cost-effective and scalable AI service paradigm, namely, ‘AI as a Service.’ We summarize current state-of-the-art solutions for AI-as-a-Service (AIaaS), and we discuss its prospects for growth and opportunities to advance the concept. To this end, we perform a thorough review of recent research on AI and various deployment strategies for emerging domains considering both technical as well as survey articles. Next, we identify various characteristics and capabilities that need to be met before an AIaaS model can be successfully designed and deployed. Based on this we present a general framework of an AIaaS architecture that integrates the required aaS characteristics with the capabilities of AI. We also compare various approaches for offering AIaaS to end users. Finally, we illustrate several real-world use cases for AIaaS models, followed by a discussion of some of the challenges that must be addressed to enable AIaaS adoption.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"39 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2025-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143072449","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}
Visual Question Answering (VQA) is a challenge task that combines natural language processing and computer vision techniques and gradually becomes a benchmark test task in multimodal large language models (MLLMs). The goal of our survey is to provide an overview of the development of VQA and a detailed description of the latest models with high timeliness. This survey gives an up-to-date synthesis of natural language understanding of images and text, as well as the knowledge reasoning module based on image-question information on the core VQA tasks. In addition, we elaborate on recent advances in extracting and fusing modal information with vision-language pretraining models and multimodal large language models in VQA. We also exhaustively review the progress of knowledge reasoning in VQA by detailing the extraction of internal knowledge and the introduction of external knowledge. Finally, we present the datasets of VQA and different evaluation metrics and discuss possible directions for future work.
{"title":"Natural Language Understanding and Inference with MLLM in Visual Question Answering: A Survey","authors":"Jiayi Kuang, Ying Shen, Jingyou Xie, Haohao Luo, Zhe Xu, Ronghao Li, Yinghui Li, Xianfeng Cheng, Xika Lin, Yu Han","doi":"10.1145/3711680","DOIUrl":"https://doi.org/10.1145/3711680","url":null,"abstract":"Visual Question Answering (VQA) is a challenge task that combines natural language processing and computer vision techniques and gradually becomes a benchmark test task in multimodal large language models (MLLMs). The goal of our survey is to provide an overview of the development of VQA and a detailed description of the latest models with high timeliness. This survey gives an up-to-date synthesis of natural language understanding of images and text, as well as the knowledge reasoning module based on image-question information on the core VQA tasks. In addition, we elaborate on recent advances in extracting and fusing modal information with vision-language pretraining models and multimodal large language models in VQA. We also exhaustively review the progress of knowledge reasoning in VQA by detailing the extraction of internal knowledge and the introduction of external knowledge. Finally, we present the datasets of VQA and different evaluation metrics and discuss possible directions for future work.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"14 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2025-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143072452","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}
Separation of Duty (SoD) is a fundamental security principle that ensures that critical tasks or functions are divided upon multiple users in order to prevent fraud. The topic of SoD spans over many different areas like Identity and Access Management, Workflows, Petri nets or high-level enterprise management. In this survey paper we conduct a systematic and stand-alone literature review on SoD. We develop a multi-level classification scheme and analyse the state of the art and current trends in SoD research as well as the current challenges and potential research gaps. To the best of our knowledge, this is the first effort to comprehensively survey and structure SoD literature.
{"title":"Separation of Duty in Information Security","authors":"Sebastian Groll, Ludwig Fuchs, Günther Pernul","doi":"10.1145/3715959","DOIUrl":"https://doi.org/10.1145/3715959","url":null,"abstract":"Separation of Duty (SoD) is a fundamental security principle that ensures that critical tasks or functions are divided upon multiple users in order to prevent fraud. The topic of SoD spans over many different areas like Identity and Access Management, Workflows, Petri nets or high-level enterprise management. In this survey paper we conduct a systematic and stand-alone literature review on SoD. We develop a multi-level classification scheme and analyse the state of the art and current trends in SoD research as well as the current challenges and potential research gaps. To the best of our knowledge, this is the first effort to comprehensively survey and structure SoD literature.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"20 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2025-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143057123","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}
Graph Neural Networks (GNNs) have received significant attention for demonstrating their capability to handle graph data. However, they are difficult to be deployed in resource-limited devices because of model sizes and scalability constraints imposed by the multi-hop data dependency. In addition, real-world graphs usually possess complex structural information and features. Therefore, to improve the applicability of GNNs and fully encode the complicated topological information, Knowledge Distillation on Graphs (KDG) has been introduced to build a smaller but effective model, leading to model compression and performance improvement. Recently, KDG has achieved considerable progress, with many studies proposed. In this survey, we systematically review these works. Specifically, we first introduce the challenges and bases of KDG, then categorize and summarize the existing work of KDG by answering the following three questions: 1) what to distillate, 2) who to whom, and 3) how to distillate. We offer in-depth comparisons and elucidate the strengths and weaknesses of each design. Finally, we share our thoughts on future research directions.
{"title":"Knowledge Distillation on Graphs: A Survey","authors":"Yijun Tian, Shichao Pei, Xiangliang Zhang, Chuxu Zhang, Nitesh Chawla","doi":"10.1145/3711121","DOIUrl":"https://doi.org/10.1145/3711121","url":null,"abstract":"Graph Neural Networks (GNNs) have received significant attention for demonstrating their capability to handle graph data. However, they are difficult to be deployed in resource-limited devices because of model sizes and scalability constraints imposed by the multi-hop data dependency. In addition, real-world graphs usually possess complex structural information and features. Therefore, to improve the applicability of GNNs and fully encode the complicated topological information, Knowledge Distillation on Graphs (KDG) has been introduced to build a smaller but effective model, leading to model compression and performance improvement. Recently, KDG has achieved considerable progress, with many studies proposed. In this survey, we systematically review these works. Specifically, we first introduce the challenges and bases of KDG, then categorize and summarize the existing work of KDG by answering the following three questions: 1) what to distillate, 2) who to whom, and 3) how to distillate. We offer in-depth comparisons and elucidate the strengths and weaknesses of each design. Finally, we share our thoughts on future research directions.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"53 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2025-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143056579","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}
Peering has rapidly emerged as a critical element of the Internet architecture in ensuring low latency, bandwidth efficient and high Quality-of-Service operation for Internet Service Providers (ISPs). In this paper, we delve into the multifaceted domain of peering, providing a comprehensive overview of its fundamental elements. We investigate the various peering models and policies available to ISPs, as well as key players of the ecosystem. We also explore security risks associated with peering and available safeguards to mitigate them. Our goal is to review current industry practices and research literature to shed valuable insights into the complexities of ISP peering.
{"title":"Resource Sharing on the Internet: A Comprehensive Survey on ISP Peering","authors":"Anindo Mahmood, Murat Yuksel","doi":"10.1145/3715906","DOIUrl":"https://doi.org/10.1145/3715906","url":null,"abstract":"Peering has rapidly emerged as a critical element of the Internet architecture in ensuring low latency, bandwidth efficient and high Quality-of-Service operation for Internet Service Providers (ISPs). In this paper, we delve into the multifaceted domain of peering, providing a comprehensive overview of its fundamental elements. We investigate the various peering models and policies available to ISPs, as well as key players of the ecosystem. We also explore security risks associated with peering and available safeguards to mitigate them. Our goal is to review current industry practices and research literature to shed valuable insights into the complexities of ISP peering.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"147 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143056630","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}
Set function learning has emerged as a crucial area in machine learning, addressing the challenge of modeling functions that take sets as inputs. Unlike traditional machine learning that involves fixed-size input vectors where the order of features matters, set function learning demands methods that are invariant to permutations of the input set, presenting a unique and complex problem. This survey provides a comprehensive overview of the current development in set function learning, covering foundational theories, key methodologies, and diverse applications. We categorize and discuss existing approaches, focusing on deep learning approaches, such as DeepSets and Set Transformer based methods, as well as other notable alternative methods beyond deep learning, offering a complete view of current models. We also introduce various applications and relevant datasets, such as point cloud processing and multi-label classification, highlighting the significant progress achieved by set function learning methods in these domains. Finally, we conclude by summarizing the current state of set function learning approaches and identifying promising future research directions, aiming to guide and inspire further advancements in this promising field.
{"title":"Advances in Set Function Learning: A Survey of Techniques and Applications","authors":"Jiahao Xie, Guangmo Tong","doi":"10.1145/3715905","DOIUrl":"https://doi.org/10.1145/3715905","url":null,"abstract":"Set function learning has emerged as a crucial area in machine learning, addressing the challenge of modeling functions that take sets as inputs. Unlike traditional machine learning that involves fixed-size input vectors where the order of features matters, set function learning demands methods that are invariant to permutations of the input set, presenting a unique and complex problem. This survey provides a comprehensive overview of the current development in set function learning, covering foundational theories, key methodologies, and diverse applications. We categorize and discuss existing approaches, focusing on deep learning approaches, such as DeepSets and Set Transformer based methods, as well as other notable alternative methods beyond deep learning, offering a complete view of current models. We also introduce various applications and relevant datasets, such as point cloud processing and multi-label classification, highlighting the significant progress achieved by set function learning methods in these domains. Finally, we conclude by summarizing the current state of set function learning approaches and identifying promising future research directions, aiming to guide and inspire further advancements in this promising field.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"53 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143056758","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}
A user-focused technological approach is essential for privacy and data protection, so a systematic mapping study was conducted to review how researchers approach such matters. Of 8867 papers, 231 were systematically selected and analysed. Through thematic analysis, we identified three main themes: improving privacy policies, raising privacy awareness, and controlling information disclosure. Notably, 45% of the studies lacked user involvement, highlighting a diverse landscape in the extent of real user participation in research evaluations. This study provides valuable insights for researchers and practitioners in promoting privacy-preserving human-computer interaction.
{"title":"User-Centred Privacy and Data Protection: An Overview of Current Research Trends and Challenges for the Human-Computer Interaction Field","authors":"Shirlei Aparecida de Chaves, Fabiane Benitti","doi":"10.1145/3715903","DOIUrl":"https://doi.org/10.1145/3715903","url":null,"abstract":"A user-focused technological approach is essential for privacy and data protection, so a systematic mapping study was conducted to review how researchers approach such matters. Of 8867 papers, 231 were systematically selected and analysed. Through thematic analysis, we identified three main themes: improving privacy policies, raising privacy awareness, and controlling information disclosure. Notably, 45% of the studies lacked user involvement, highlighting a diverse landscape in the extent of real user participation in research evaluations. This study provides valuable insights for researchers and practitioners in promoting privacy-preserving human-computer interaction.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"36 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143057124","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}
Dinh-Viet-Toan Le, Louis Bigo, Dorien Herremans, Mikaela Keller
– Music is frequently associated with the notion of language as both domains share several similarities, including the ability for their content to be represented as sequences of symbols. In computer science, the fields of Natural Language Processing (NLP) and Music Information Retrieval (MIR) reflect this analogy through a variety of similar tasks, such as author detection or content generation. This similarity has long encouraged the adaptation of NLP methods to process musical data, in particular symbolic music data, and the rise of Transformer neural networks has considerably strengthened this practice. This survey reviews NLP methods applied to symbolic music generation and information retrieval following two axes. We first propose an overview of representations of symbolic music inspired by text sequential representations. We then review a large set of computational models, in particular deep learning models, that have been adapted from NLP to process these musical representations for various MIR tasks. These models are described and categorized through different prisms with a highlight on their music-specialized mechanisms. We finally present a discussion surrounding the adequate use of NLP tools to process symbolic music data. This includes technical issues regarding NLP methods which may open several doors for further research into more effectively adapting NLP tools to symbolic MIR.
{"title":"Natural Language Processing Methods for Symbolic Music Generation and Information Retrieval: A Survey","authors":"Dinh-Viet-Toan Le, Louis Bigo, Dorien Herremans, Mikaela Keller","doi":"10.1145/3714457","DOIUrl":"https://doi.org/10.1145/3714457","url":null,"abstract":"– Music is frequently associated with the notion of language as both domains share several similarities, including the ability for their content to be represented as sequences of symbols. In computer science, the fields of Natural Language Processing (NLP) and Music Information Retrieval (MIR) reflect this analogy through a variety of similar tasks, such as author detection or content generation. This similarity has long encouraged the adaptation of NLP methods to process musical data, in particular symbolic music data, and the rise of Transformer neural networks has considerably strengthened this practice. This survey reviews NLP methods applied to symbolic music generation and information retrieval following two axes. We first propose an overview of representations of symbolic music inspired by text sequential representations. We then review a large set of computational models, in particular deep learning models, that have been adapted from NLP to process these musical representations for various MIR tasks. These models are described and categorized through different prisms with a highlight on their music-specialized mechanisms. We finally present a discussion surrounding the adequate use of NLP tools to process symbolic music data. This includes technical issues regarding NLP methods which may open several doors for further research into more effectively adapting NLP tools to symbolic MIR.","PeriodicalId":50926,"journal":{"name":"ACM Computing Surveys","volume":"36 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2025-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143055593","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：