Page 368 - 《软件学报》2026年第2期

姬涛 等: AI 赋能的关系型数据库系统研究: 标准化、技术与挑战 847

习的联合动作空间; 研发轻量化嵌入式模型, 减少 GPU 内存占用.
数据库开发接口需构建标准化接口框架, 支持动态模型适配、安全数据管道、跨平台部署, 同时提供可视化
调试工具, 降低 AI 组件集成门槛.
尽管当前智能数据库的各个研究领域已初步建立起标准化流程框架, 但是, 面对人工智能技术的动态演进与
复杂应用场景的膨胀, 突破现有框架已成为必然. 这主要因为: 标准化框架预设的静态模块隔离规则难以兼容 AI
驱动的技术现实. 未来智能数据库的发展方向可能会集中于从“数据库+AI 插件”的松散耦合转向 AI 原生设计. 数
据库社区需要开发便捷查询、易于维护和快速响应的完全由人工智能驱动的数据库. 人工智能的技术现阶段需要
提升其在数据库系统中的适用性和准确性, 以更好地应对动态变化的数据和查询. 包括开发更准确高效的模型、
深化自然语言理解能力、增强实时数据处理能力, 以及在分布式和云数据库环境中优化数据同步和处理能力. 数
据库的智能化将成为未来数据库发展的必然趋势, 智能数据库也会为日益增长的应用数据需求提供更加强大和有
效的解决方案.

References
[1] Sun LM, Zhang SM, Ji T, Li CP, Chen H. Survey of data management techniques powered by artificial intelligence. Ruan Jian Xue
Bao/Journal of Software, 2020, 31(3): 600–619 (in Chinese with English abstract). http://www.jos.org.cn/1000-9825/5909.htm [doi: 10.
13328/j.cnki.jos.005909]
[2] Li GL, Zhou XH, Cao L. AI meets database: AI4DB and DB4AI. In: Proc. of the 2021 Int’l Conf. on Management of Data. ACM, 2021.
2859–2866. [doi: 10.1145/3448016.3457542]
[3] Ding JL, Marcus R, Kipf A, Nathan V, Nrusimha A, Vaidya K, van Renen A, Kraska T. SageDB: An instance-optimized data analytics
system. Proc. of the VLDB Endowment, 2022, 15(13): 4062–4078. [doi: 10.14778/3565838.3565857]
[4] Kraska T, Alizadeh M, Beutel A, Chi EH, Kristo A, Leclerc G, Madden S, Mao HZ, Nathan V. SageDB: A learned database system.
2019. https://www.cidrdb.org/cidr2019/papers/p117-kraska-cidr19.pdf
[5] Ma L, Zhang W, Jiao J, Wang WW, Butrovich M, Lim WS, Menon P, Pavlo A. MB2: Decomposed behavior modeling for self-driving
database management systems. In: Proc. of the 2021 Int’l Conf. on Management of Data. ACM, 2021. 1248–1261. [doi: 10.1145/
3448016.3457276]
[6] Li GL, Zhou XH, Sun J, Yu X, Han Y, Jin LY, Li WB, Wang TQ, Li SF. OpenGauss: An autonomous database system. Proc. of the
VLDB Endowment, 2021, 14(12): 3028–3042. [doi: 10.14778/3476311.3476380]
[7] Zhao ZH, Cai SF, Gao HT, Pan HX, Xiang SQ, Xing NL, Chen G, Ooi BC, Shen YY, Wu YC, Zhang MH. NeurDB: On the design and
implementation of an AI-powered autonomous database. arXiv:2408.03013, 2025.
[8] Helskyaho H, Yu J, Yu K. Machine Learning for Oracle Database Professionals: Deploying Model-driven Applications and Automation
Pipelines. Berkeley: Apress, 2021. [doi: 10.1007/978-1-4842-7032-5]
[9] Barnes J. Microsoft Azure Essentials Azure Machine Learning. Washington: Microsoft Press, 2015.
[10] Narula S, Jain A, Prachi. Cloud computing security: Amazon Web service. In: Proc. of the 5th Int’l Conf. on Advanced Computing &
Communication Technologies. Haryana: IEEE, 2015. 501–505. [doi: 10.1109/ACCT.2015.20]
[11] Verbitski A, Gupta A, Saha D, Brahmadesam M, Gupta K, Mittal R, Krishnamurthy S, Maurice S, Kharatishvili T, Bao XF. Amazon
Aurora: Design considerations for high throughput cloud-native relational databases. In: Proc. of the 2017 ACM Int’l Conf. on
Management of Data. Chicago: ACM, 2017. 1041–1052. [doi: 10.1145/3035918.3056101]
[12] Färber F, Cha SK, Primsch J, Bornhövd C, Sigg S, Lehner W. SAP HANA database: Data management for modern business
applications. ACM SIGMOD Record, 2012, 40(4): 45–51. [doi: 10.1145/2094114.2094126]
[13] Gupta A, Tyagi S, Panwar N, Sachdeva S, Saxena U. NoSQL databases: Critical analysis and comparison. In: Proc. of the 2017 Int’l
Conf. on Computing and Communication Technologies for Smart Nation. Gurgaon: IEEE, 2017. 293–299. [doi: 10.1109/IC3TSN.2017.
8284494]
[14] Bhattacharyya A, Chakravarty D. Graph database: A survey. In: Proc. of the 2020 Int’l Conf. on Computer, Electrical & Communication
Engineering (ICCECE). Kolkata: IEEE, 2020. 1–8. [doi: 10.1109/ICCECE48148.2020.9223105]
[15] Zhong V, Xiong CM, Socher R. Seq2SQL: Generating structured queries from natural language using reinforcement learning.
arXiv:1709.00103, 2017.
[16] Xu XJ, Liu C, Song D. SQLNet: Generating structured queries from natural language without reinforcement learning.
arXiv:1711.04436, 2017.