人机结合的智能情报分析系统设计与实现
Design and Implementation of an Intelligent Human-Machine Intelligence Analysis System
丁洪鑫 1汪榕 1周维 1曹扬 1支婷 1董厚泽1
作者信息
- 1. 中电科大数据研究院有限公司,贵州 贵阳 550022;提升政府治理能力大数据应用技术国家工程研究中心,贵州 贵阳 550022
- 折叠
摘要
文章首先对情报分析方法与人工智能的融合方式进行辨析,然后以AI赋能、人在回路的情报分析过程作为主导设计思想对智能情报分析系统进行整体功能架构及数据架构设计,同时为了降低使用门槛,提升数据分析结果的可解释性和准确性,提出了大语言模型与知识图谱的组合架构,并开发了人机结合的智能情报分析系统.该平台具备便于集成机器学习算法的架构,采用面向非技术用户的自然语言交互接口,使依赖人类智能的情报分析方法与依赖人工智能的大数据分析方法有机结合,极大地增强了分析人员的数据融合分析能力.
Abstract
This paper first analyzes the integration of intelligence analysis methods and artificial intelligence,and then takes AI empowerment and human in the loop intelligence analysis process as the main design concept to design the overall functional architecture and data architecture of the intelligent intelligence analysis system.At the same time,in order to reduce the threshold for use,improve the interpretability and accuracy of data analysis results,a combination architecture of large language models and Knowledge Graphs is proposed,and an intelligent human-machine intelligence analysis system is developed.This platform has an architecture that facilitates the integration of machine learning algorithms.It adopts a natural language interaction interface for non-technical users,which organically combines intelligence analysis methods that rely on human intelligence with big data analysis methods that rely on artificial intelligence,greatly enhancing the data fusion analysis ability of analysts.
关键词
情报分析/数据融合分析/大语言模型/知识图谱/可视化分析工具Key words
intelligence analysis/data fusion analysis/large language model/Knowledge Graph/visual analysis tool引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据