[1]王浩田,鄂海红,王 勇,等.一种云原生数据智能服务生产平台设计与实现[J].计算机技术与发展,2023,33(10):15-21.[doi:10. 3969 / j. issn. 1673-629X. 2023. 10. 003]
　WANG Hao-tian,E Hai-hong,WANG Yong,et al.Design and Implementation of a Cloud Native Data Intelligent Service Production Platform[J].,2023,33(10):15-21.[doi:10. 3969 / j. issn. 1673-629X. 2023. 10. 003]

点击复制

一种云原生数据智能服务生产平台设计与实现()

分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:

33

期数:

2023年10期

页码:

15-21

栏目:

分布与并行计算

出版日期:

2023-10-10

文章信息/Info

Title:

Design and Implementation of a Cloud Native Data Intelligent Service Production Platform

文章编号:

1673-629X(2023)10-0015-07

作者:

王浩田¹ ; 鄂海红¹ ; 王 勇² ; 宋美娜¹

1. 北京邮电大学 计算机学院(国家示范性软件学院),北京 100876;

2. 联洋国融(北京)科技有限公司,北京 100088

Author(s):

WANG Hao-tian1 ; E Hai-hong1 ; WANG Yong2 ; SONG Mei-na1

1. School of Computer Science ( National Pilot Software Engineering School) ,Beijing University of Posts and Telecommunications,Beijing 100876,China;

2. Lianyang Guorong ( Beijing) Technology Co. ,Ltd. , Beijing 100088,China

关键词:

数据智能; 云原生; 大数据; 人工智能; 持续生产

Keywords:

data intelligence; cloud native; big data; artificial intelligence; continuous production

分类号:

TP311

DOI:

10. 3969 / j. issn. 1673-629X. 2023. 10. 003

摘要:

将数据智能服务嵌入到实际业务场景中创造价值,已成为众多企业数字化转型的重要特征。 针对数据智能服务落地因过程多且复杂而导致的开发周期长、成本高等问题,设计并实现了一种数据智能服务生产平台。 该平台底层基于Kubernetes 容器云并以容器的方式对外提供了计算、存储、虚拟化和监控等基础服务,中间层基于这些基础服务搭建了用于海量多源异构数据统一组织管理的数据仓库、用于自动化生产数据智能服务的生产线和用于细粒度复用已有服务生产新服务的服务引擎等基础设施,顶层按照不同的业务场景对这些基础设施提供的能力进行了封装并设计了数据管理与标注、算法模型研发、模型批量生产和数据智能服务管理等模块。 实际运行结果表明,该平台有效提高了数据资产、模型资产、服务资产之间相互转变的自动化能力,缩短了服务开发周期、提高了服务开发效率并降低了人力成本的投入。

Abstract:

Embedding data intelligence service into actual business scenarios to create value has become an important feature of many enterprises’ digital transformation.?

Aiming at the problems of long development cycle and high cost caused by the multiple and complexprocess of data intelligence service landing,we design and implement a data intelligence service production platform. The bottom layer ofthe platform is based on the Kubernetes container cloud and provides external basic services such as computing,storage,virtualization andmonitoring in the form of containers. Based on these basic services, the middle layer has built a data warehouse for the unifiedorganization and management of massive multi-source heterogeneous data,a production line for automated production of data intelligentservices,and a service engine for fine - grained reuse of existing services to produce new services. The top layer encapsulates thecapabilities provided by these infrastructures according to different business scenarios and designs modules such as data management andannotation,algorithm model development,model mass production and data intelligent service management. The actual operation resultsshow that the platform effectively improves the automation ability of the mutual transformation among data assets, model assets andservice assets,shortens the service development cycle,improves the service development efficiency and reduces the input of labor cost.

常用功能