情报杂志2025，Vol.44Issue(5)：191-198,8.DOI:10.3969/j.issn.1002-1965.2025.05.023
联邦学习赋能数智化科技情报工作中的多源数据融合
Empowering Intelligent Technology Information Work with Federated Learning for Multi-Source Data Integration
摘要
Abstract
[Research purpose]In the context of data-driven intelligence,federated learning offers an innovative solution to the issues of data silos,privacy breaches,and computational resource constraints associated with traditional machine learning methods.By analyzing the theoretical foundation and practical applications of federated learning in the integration of multi-source data for technology information,this study aims to promote the transformation and upgrading of multi-source data integration in technology information,achieving data sha-ring and secure integration,and enhancing the precision and security of technology information analysis.[Research method]By combi-ning the technical characteristics of federated learning with the practical points of multi-source data integration in technology information work,it analyzes the theoretical foundation of federated learning in multi-source data integration in technology information from three as-pects:information confidentiality,information optimization,and information collaboration.It then explores the application scenarios of federated learning empowering multi-source data integration in technology information from four dimensions:cross-organization,cross-region,cross-domain,and cross-volume.Subsequently,it proposes specific paths to promote multi-source data integration in technology information from the perspectives of technological innovation,multiple models,comprehensive perception,and multi-party collaboration.[Research result/conclusion]Through the application of federated learning technology,technology information work can achieve effi-cient integration and sharing of multi-source data while protecting data privacy and ensuring data security,thereby promoting the continu-ous optimization and innovative development of technology information analysis models.关键词
数智驱动/联邦学习/科技情报/多源数据融合/安全共享/数据安全Key words
data-driven intelligence/federated learning/technology information/multi-source data integration/secured sharing/data se-curity引用本文复制引用
窦路遥,魏凤,邓阿妹,周洪,辛竹琳,付豪..联邦学习赋能数智化科技情报工作中的多源数据融合[J].情报杂志,2025,44(5):191-198,8.基金项目
国家电网公司总部科技项目"双碳目标下电力绿色低碳关键支撑技术评价方法和专利标准化研究"(编号:1400-202340338A-1-1-ZN)研究成果. （编号:1400-202340338A-1-1-ZN）
