摘要
近年来随着财政体制改革浪潮的深入推进,财政信息化的建设水平也得到了大幅提升,各地陆续部署并实施了涵盖收入、预算编审、预算执行等20余项核心业务的应用系统。虽然这些系统为提高财政管理规范化、科学化、精细化水平提供了有力支撑,但不可否认系统在架构设计方面普遍存在着数据集成方式单一,决策冲突消解困难等问题。为此,论文提出了一种基于分层结构的财政综合分析与智能决策支持系统架构模型。该模型通过自动数据抽取,转换和加载(Extraction,Transformation and Loading,ETL)系统核心业务数据后建立财政数据中心;并按照混合决策Agent关心主题建立不同的知识模型,采用仲裁模型消除决策冲突,支持混合智能体对不同兴趣点的协商决策。旨在以科学的系统架构助推地方财政管理科学化、规范化进程,提升财政管理效能。通过从稳定性、准确性及适用性初步分析与对比结果来看,论文为地方财政综合分析与群体智能决策支持系统设计的架构模型具有良好的性能稳定性、决策能力及普适性。
In recent years,with the trend of financial system reform,the construction of financial informatization has also been further promoted.All regions have successively deployed and implemented application systems covering more than 20 core businesses,such as revenue,budget compilation,review and budget implementation.Although these systems provide strong sup⁃port for improving the standardization,scientization and refinement of financial management,it is undeniable that there are some de⁃fects in the system architecture design,such as the single way of data integration and the difficulty of resolving decision conflicts.Therefore,this paper designs an architecture model of financial comprehensive analysis and group intelligent decision support sys⁃tem based on knowledge representation and hybrid agent reasoning and decision-making.The model establishes a financial data cen⁃ter after automatic data extraction,transformation and loading(ETL)of the system core business data.Different knowledge models are established according to the topics concerned by the hybrid decision agent,and the arbitration model is used to eliminate the de⁃cision conflict,so as to support the negotiation and decision-making of the hybrid agent on different points of interest.It aims to pro⁃mote the scientific and standardized process of local financial management and improve the efficiency of financial management with a scientific system architecture.The stability,decision-making effect and preliminary comparison results of the applicability show that the architecture model designed for the local financial comprehensive analysis and swarm intelligence decision support system has satisfactory stability,decision-making ability and adaptability.
作者
谢生龙
崔祥
刘潇
溥颖
XIE Shenglong;CUI Xiang;LIU Xiao;PU Ying(College of Mathematics and Computer Science,Yan'an University,Yan'an 716000;Zichang Municipal Finance Bureau,Zichang 717300;School of Computer Science and Technology,Xidian University,Xi'an 710071)
出处
《计算机与数字工程》
2023年第1期1-7,124,共8页
Computer & Digital Engineering
基金
国家自然科学基金项目(编号:62262067)
陕西省自然科学基础研究计划项目(编号:2023-JC-QN-0744)
延安大学博士科研启动项目(编号:YDBK2018-35)
延安大学重点项目(编号:YDZ2019-04)
陕西省大学生创新创业训练计划项目(编号:S202110719083)资助
关键词
财政信息化
综合分析
智能决策
决策支持
financial informatization
comprehensive analysis
intelligent decision-making
decision support
