首页|一种基于大模型的数据协同服务平台构建技术

一种基于大模型的数据协同服务平台构建技术

扫码查看
随着智能服务的发展,越来越多的数据被整合起来,形成了超级复杂的服务系统,系统中各类应用产生的数据规模大、类型多、产生速度快、语义杂,导致数据的协同异常困难.大模型具有强大的理解能力及生成能力,提供了一种复杂语义贯通的跨库表协同赋能模式,有助于数据间高效交互.本文提出了一种新颖的基于大模型的数据协同服务机制,并通过实验证明其能够有效提升平台中各类数据的交互性能.
A Technology for Building a Data Collaborative Service Platform Based on Large Models
With the development of intelligent services,more and more data is being integrated to form super complex service systems.The data generated by various applications in the system is large in scale,diverse in types,fast in generation speed,and semantically complex,which makes data collaboration extremely difficult.Large models have strong understanding and generation capabilities,providing a cross database table collaborative empowerment mode with complex semantic connectivity,which facilitates efficient interaction between data.This article proposes a novel data collaboration service mechanism based on large models,and proves through experiments that it can effectively improve the interaction performance of various types of data in the platform.

multimodal dataknowledge alignmentknowledge completionlink predictionknowledge graph

朱思颖、张林超、周树德

展开 >

中国电子科技集团公司信息科学研究院,北京 100042

多模态数据 知识对齐 知识补全 链接预测 知识图谱

2024

软件
中国电子学会 天津电子学会

软件

影响因子:1.51
ISSN:1003-6970
年,卷(期):2024.45(8)