面向主题领域的科技情报智能感知系统构建研究

Research on the Construction of Subject-oriented Scientific and Technical Information Intelligent Perception System

[目的/意义]融合大变局背景下科技情报能力新需求与智能技术发展新特征,探索情报感知能力智能化发展路径。[方法/过程]深入研究主题建模、加权分类技术,实现从千万级开源情报数据中精准识别、追踪、监测特定主题领域的知识,同时通过自然语言处理、知识图谱技术,实现针对识别到的知识持续进行实体、关系抽取并自动化构建领域知识图谱,提供态势揭示、图谱探索、关系推理等知识感知能力。[结果/结论]依托系统构建的智能空战等主题领域知识库,实现了针对关注对象的持续自主扫描、监测、跟踪与实体、关系挖掘,形成了情报感知能力智能化发展路径。

[Purpose/significance]Combining the new requirements of scientific and technical information capability with the new characteristics of intelligent technology development under the background of great changes,the intelligent development path of information perception capability is explored.[Method/process]This research further studies topic modeling and weighted classification technology,and realizes accurate recognition,tracking and monitoring of knowledge in specific subject areas from tens of millions of open source information data.Meanwhile,through natural language processing and knowledge graph technology,entity and relationship extraction can be carried out for the recognized knowledge continuously,and knowledge graphs of domains can be automatically constructed to provide knowledge perception capabilities such as situation reveal,graph exploration and relationship reasoning.[Result/conclusion]Relying on the repository of intelligent air combat and other subject areas constructed by the system,it realizes continuously autonomous scanning,monitoring,tracking as well as entity and relationship mining for the concerned objects,forming the intelligent development path of information perceptibility.