Foundations and Trends in Databases

Foundations and Trends in Databases publishes one book-length survey per issue, each written by invitation and running to roughly a hundred pages, a format closer to the research monograph than to the research article.《数据库基础与趋势》每期只刊一篇书长综述，均系约稿，篇幅约一百页，形制更接近研究专著而非研究论文。

Now Publishers of Hanover, Massachusetts, opened the title in 2007, and Emerald took the publishing side over at volume 16 in 2026, keeping the ISSN and the single-article issue intact.该刊 2007 年由马萨诸塞州汉诺威的 Now Publishers 创办，2026 年第 16 卷起出版事务转由 Emerald 承接，刊号与「一期一文」的形制都没有变。

Its remit is database systems research, and the 2025 Journal Impact Factor of 1.9 puts it in the third quartile at 78 of 128 titles in computer science, software engineering.选题落在数据库系统研究，2025 年的期刊影响因子为 1.9，在计算机科学软件工程类目的 128 种刊物中列第 78 位，属第三区。

Scope

• Query processing and optimization run through the catalogue, the opening monograph of 2007 taking adaptive query processing as its subject and later items returning to how optimizers are extended in practice.查询处理与优化贯穿全目，2007 年的开篇专著以自适应查询处理为题，此后的篇目回到优化器在实践中的扩展方式。^[1]

• Storage engines and index structures make up a second block, with one monograph on B-tree technique in 2011, a sequel in 2024, column-oriented systems in 2013 and main-memory databases in 2017.存储引擎与索引结构构成第二块，B 树技术的专著 2011 年一部、2024 年又有续篇，2013 年的列式系统与 2017 年的主存数据库同属此列。

• Data integration and knowledge construction are covered at length, from information extraction in 2008 to a 2021 monograph on building and curating large knowledge bases, with Datalog engines and natural language interfaces following.数据集成与知识构建占了相当篇幅，从 2008 年的信息抽取，到 2021 年一部讲大型知识库构建与整理的专著，随后又有 Datalog 引擎与自然语言接口。

• Provenance has drawn two monographs, one in 2009 on why, how and where it is recorded and another in 2021, and a 2021 item extended the question to explaining the output of data-driven systems.数据溯源先后有两部专著，2009 年一部讨论记录什么、如何记录与记在何处，2021 年一部作了系统梳理，同年另有专著把问题延伸到对数据驱动系统输出的解释。^[2]

• Approximate answers form a thread of their own, with synopses for massive data in 2011, approximate string processing in the same period and query processing over probabilistic data in 2017.近似答案自成一条线索，2011 年有大规模数据的概要结构，同期有近似字符串处理，2017 年则有概率数据上的查询处理。

• Distributed and cloud data management is a large part of the list, covering cloud services in 2021, massively parallel processing and MapReduce in 2013, and the algorithmic side of parallel data processing in 2018.分布式与云数据管理在目录中占比不小，2021 年的云服务、2013 年的大规模并行处理与 MapReduce，以及 2018 年并行数据处理的算法层面都在其内。

• Consensus and replication have a run of their own, from a 2023 monograph on distributed commit and blockchain to a 2026 treatment of state machine replication in distributed databases.共识与复制另有一条线索，从 2023 年一部论述分布式提交与区块链的专著，到 2026 年一部论述分布式数据库中状态机复制的长篇。

• Machine learning enters from two directions, with a 2020 monograph on distributed learning systems built from first-order methods and a 2024 survey of learned query optimizers.机器学习从两个方向进入，2020 年有基于一阶方法的分布式学习系统专著，2024 年有学习型查询优化器的梳理。

• Spatial and graph data have three entries, big spatial data in 2016, graph analytics platforms in 2017 and a 2024 monograph on querying graph-structured databases.空间与图数据有三部，2016 年的大空间数据，2017 年的图分析平台，2024 年又有一部论述图结构数据库查询的专著。

• Privacy and security appear early and late, with privacy-preserving data publishing in 2009, access control and secure aggregation in 2011, and differential privacy for databases in 2021.隐私与安全前后都有落点，2009 年有隐私保护数据发布，2011 年有访问控制与安全聚合，2021 年则有面向数据库的差分隐私。

• Hardware acceleration reached the list through two monographs, one on FPGA-accelerated analytics in 2020 and one on database systems on GPUs in 2021.硬件加速通过两部专著进入目录，2020 年一部讲 FPGA 加速的分析处理，2021 年一部讲运行在 GPU 上的数据库系统。

• Data quality is treated once, in a 2015 monograph on consistency and deduplication in relational data cleaning, a subject the list has not returned to since.数据质量只被处理过一次，即 2015 年那部论述关系数据清洗中一致性与去重的专著，此后未再涉及。

• Applied work closes the range, with medical data infrastructure in 2017, a review of visualization recommendation systems in 2024 and analytical queries over unstructured data in 2025.应用类篇目收束整个范围，2017 年有医学数据基础设施，2024 年有可视化推荐系统的系统梳理，2025 年有面向非结构化数据的分析查询。

History

• Volume 1, number 1 appeared on 8 August 2007 as a monograph on adaptive query processing running to 140 pages, and the ISSN 1931-7883 was registered to a publisher in the United States.第 1 卷第 1 期于 2007 年 8 月 8 日面世，是一部 140 页的自适应查询处理专著，刊号 1931-7883 当时登记在美国的出版方名下。

• The founder was Now Publishers, a house in Hanover, Massachusetts, that built its list around the Foundations and Trends formula of short books issued as journal numbers.创办方是位于马萨诸塞州汉诺威的 Now Publishers，这家出版机构以「基础与趋势」的成套丛书立身，把篇幅不长的专著编成期刊的期号发行。

• Pagination has been continuous inside each volume from the beginning, volume 1 running through pages 1 to 140, then 141 to 259, then 261 to 377 and finally 379 to 474.每卷内部的页码自创刊起就连排，第 1 卷的四篇依次占 1 至 140 页，接着 141 至 259 页，再到 261 至 377 页，最后是 379 至 474 页。

• Double issues arrived as the monographs grew, with numbers 1 and 2 of volume 2 combined in 2009 for a 167-page item and numbers 1 to 3 of volume 4 combined in 2011 for 294 pages.专著变长之后出现了合刊，2009 年第 2 卷第 1、2 期合成一部 167 页的篇目，2011 年第 4 卷第 1 至 3 期合成 294 页的一部。

• The schedule has never been even, and two calendar years carry no record at all, 2014 and 2019, while volume 5 ran from 2013 to 2015 and volume 9 from 2020 to 2021.出版节奏从未均匀，2014 年与 2019 年两年完全没有记录，第 5 卷从 2013 年拖到 2015 年，第 9 卷从 2020 年延至 2021 年。

• By 2026 the register holds 44 monographs across 16 volumes, with a median length of 114 pages, a shortest of 53 and a longest of 383.到 2026 年，登记在册的专著共 44 部，分属 16 卷，篇幅中位数 114 页，最短 53 页，最长 383 页。

• The publisher changed hands at volume 16, whose first number appeared in April 2026 under the DOI prefix 10.1108 that belongs to Emerald, where every earlier volume sits under 10.1561.出版方在第 16 卷易手，这一卷的第 1 期于 2026 年 4 月面世，DOI 前缀为 Emerald 的 10.1108，此前各卷一律挂在 10.1561 之下。^[3]

• Crossref opened a record for the title under Emerald in December 2025, and the old series page now points at the Emerald edition of the same title.Crossref 于 2025 年 12 月为这份刊物在 Emerald 名下开出新档，同一时期旧站入口开始指向 Emerald 上的同名刊物页。

• The Foundations and Trends list is marketed as a family of about 30 titles, each issue a single commissioned review of 50 to 200 pages.「基础与趋势」这一族被推介为约 30 种刊物，每期一篇约稿综述，篇幅在 50 至 200 页之间。

• The title sits in the Emerging Sources Citation Index, ranks third quartile at 78 of 128 in computer science, software engineering, and the 2025 register counts a single citable item.该刊收录于新兴来源引文索引，在计算机科学软件工程类目列 128 种中的第 78 位，属第三区，2025 年的统计只记 1 篇可引篇目。

• That count matches the catalogue, since 2025 produced exactly one monograph, the second number of volume 15, and the citation half-life stands at 13.3 years.这一数字与目录相符，2025 年恰好只出一部专著，即第 15 卷第 2 期，被引半衰期则长达 13.3 年。

• Impact factors have moved without a trend, 2.7 in 2023, 1.0 in 2024 and 1.9 in 2025, on a total of 149 citations to the whole backlist.影响因子没有稳定走向，2023 年为 2.7，2024 年降到 1.0，2025 年回到 1.9，全部过刊累计被引 149 次。

• No article processing charge supports the list, gold open access stands at zero, and each monograph is sold on its own as a printed booklet or an ebook.这份刊物不靠文章处理费维持，金色开放获取比例为 0，每部专著单独售卖，纸本小册与电子书各成一件。

• The subject mix has drifted from engine internals toward learning and privacy, since the 2007 to 2013 volumes stay on B-trees and column stores while the 2020 to 2026 volumes take up differential privacy and replication.选题的重心从引擎内部结构移向学习与隐私，2007 至 2013 年的卷册集中在 B 树与列式存储，2020 至 2026 年的卷册则处理差分隐私与复制协议。

Editors

• The series works by commission, the subject and the author being settled by the title's editor and its board before a line is drafted, so the catalogue is a record of editorial choices.这套刊物的运作靠约稿，选题与作者由该刊主编与编委会在动笔之前定下，目录本身就是编辑抉择的记录。

• Each Foundations and Trends title carries an editor-in-chief and an editorial board drawn from the research community, and the database title has been run this way since its first volume.「基础与趋势」的每一种刊物都设主编与编委会，成员从研究群体中选任，数据库这一种自首卷起即按此办理。

• Review is by peers before acceptance, and the publisher describes the whole list as peer reviewed, with the manuscript read as a book rather than as a paper of ordinary length.录用之前须经同行评审，出版方把整族刊物都标为同行评审，审读的对象是一部书稿而非寻常篇幅的论文。

• The authors are usually the people who built what they describe, Hellerstein, Stonebraker and Hamilton writing the 2007 architecture monograph and Graefe writing both B-tree items.作者通常是其所写对象的建造者，2007 年的体系结构专著出自 Hellerstein、Stonebraker 与 Hamilton 之手，两部 B 树专著均出自 Graefe。

• A page budget is set with the author, and the finished items run from 53 to 383 pages with a median of 114, which is what separates the format from a long conference survey.篇幅与作者事先商定，成稿从 53 页到 383 页不等，中位数 114 页，这一区间把它与会议长综述区分开来。

• The register is deliberately tutorial, and the series is promoted for textbook clarity alongside journal rigour, a combination aimed at graduate teaching.行文刻意带上教程的气质，这套丛书被推介为兼具教科书般的清晰与期刊的严谨，用意在于研究生教学。

• One volume can carry four unrelated subjects, and volume 11 did exactly that in 2021 with GPU systems, differential privacy, explanations and natural language interfaces in four consecutive numbers.一卷之内可以并置四个不相干的题目，2021 年的第 11 卷正是如此，连续四期依次是 GPU 系统、差分隐私与解释，最后一期谈自然语言接口。

• Production moved with the publisher, and the Emerald office now handles the series, while volume numbering and issue pagination carried straight on, volume 16 opening again at page 1.制作随出版方一并转移，如今由 Emerald 的编辑部门经手，卷号与期内页码则原样延续，第 16 卷又从第 1 页起算。

• A proposal reaches the editor before any manuscript exists, and the board decides whether the subject belongs in the list, which keeps the catalogue to topics it can defend.先到编辑手上的是选题提案，成稿尚不存在，编委会判断这个题目是否该进这份目录，能留下来的都是它愿意担保的方向。

• The editorial stake is therefore canonical, since each monograph is meant to be the first thing a newcomer to a subfield reads and the reference older hands cite.编辑的分量因此落在定典上，每部专著都被指望成为新入某个子方向者先读的一本，也是老手引用的依据。