基于开源科技项目数据的多模态知识图谱构建研究  

Research on the Construction of Multi-Modal Knowledge Graph Based on Open Source Science and Technology Project Data

在线阅读下载全文

作  者:窦永香[1] 解哲辉 汤晓芳 Dou Yongxiang;Xie Zhehui;Tang Xiaofang(School of Economics and Management,Xidian University,Shaanxi Xi’an 710126)

机构地区:[1]西安电子科技大学经济与管理学院,陕西西安710126

出  处:《情报理论与实践》2025年第3期32-40,共9页Information Studies:Theory & Application

摘  要:[目的/意义]在全球科技竞争日益激烈的背景下,开源情报作为研究和预测世界科技发展趋势的重要工具,对于科技情报工作的重要性愈发凸显。科技项目数据作为科技情报的重要数据来源和分析对象,呈现出多源、多模态的特性,对其进行深度分析和挖掘,并服务于管理与决策活动,是新时期科技资源高效利用的重要途径。[方法/过程]文章将情报工作流程与多模态知识图谱构建流程相融合,提出了基于开源科技项目数据的多模态知识图谱构建框架,包含需求分析、开源情报采集、多模态科技项目知识图谱构建以及知识图谱服务4个关键环节,并详细讨论了每个环节中的核心内容。[结果/结论]以美国情报高级研究计划局(IARPA)公开项目的多模态数据为例,遵循上述框架构建了包含项目、机构、人员、技术、文档、图像、视频7类实体及多种关系的多模态知识图谱。未来,可将多模态知识图谱与问答系统、推荐系统相结合,为科技管理决策提供更加智能化的情报服务。[Purpose/significance]Against the backdrop of increasingly intense global technological competition,open source intelligence has become a crucial tool for researching and predicting global trends in science and technology,highlighting its growing importance in scientific and technological intelligence work.As a significant data source and analytical target for scientific and technological intelligence,science and technology project data exhibit multi-source and multi-modal characteristics.Conducting indepth analysis and mining of such data to support management and decision-making activities represents an essential approach to the efficient utilization of scientific and technological resources in the new era.[Method/process]This paper integrates the intelligence workflow with the multi-modal knowledge graph construction process,proposing a framework for constructing multi-modal knowledge graphs based on open source science and technology project data.The framework comprises four key components:demand analysis,open source intelligence collection,multi-modal science and technology project knowledge graph construction,and knowledge graph services.Each component’s core elements are discussed in detail.[Result/conclusion]Using the multi-modal data from publicly available projects of the Intelligence Advanced Research Projects Activity(IARPA)as a case study,a multi-modal knowledge graph was constructed following the proposed framework.The graph includes seven types of entities—projects,institutions,personnel,technologies,documents,images,and videos—along with various relationships.In the future,multi-modal knowledge graphs can be integrated with question-answering systems and recommendation systems to provide more intelligent intelligence services for science and technology management and decision-making.

关 键 词:开源情报 科技项目 多模态知识图谱 科技情报 数据挖掘 科技管理 

分 类 号:TP391.1[自动化与计算机技术—计算机应用技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象