基于CBR的应急情报智能决策支持系统研究

发布时间：2018-03-04 14:11

本文选题：应急情报　切入点：应急决策　出处：《情报杂志》2017年10期 　论文类型：期刊论文

【摘要】：[目的/意义]针对应急情报智能决策支持系统(IDSS)自学习能力较差等问题,提出了一种基于案例推理的应急情报智能决策支持系统。[方法/过程]首先,分析了应急情报IDSS的特点和局限性;其次,分析了基于案例推理(CBR)系统和贝叶斯信念网(BBN)在应急情报智能决策中的应用;再次,使用人工神经网络(ANN)算法代替CBR系统中原有检索算法,使用BBN训练分类器,建立了基于ANN-BBN-CBR的应急情报IDSS模型;最后,给出了该系统的工作流程,并通过收集到的数据验证了该系统有效性和可行性。[结果/结论]该系统模型的使用不仅使大型案例库的检索和匹配速度及准确率得到了大幅度的提高,并且为案例相似性的度量提供了适当的统计信息,给案例库的组织管理也带来了方便,更为应急情报收集、整合、分析、传递、反馈等工作提供了研究思路。
[Abstract]:[objective / significance] in view of the poor self-learning ability of IDSS, an intelligent decision support system for emergency intelligence based on case-based reasoning (CBR) is proposed. [method / process] first of all, The characteristics and limitations of emergency intelligence IDSS are analyzed. Secondly, the application of CBR-based reasoning system and Bayesian belief net in intelligent decision-making of emergency intelligence is analyzed. The artificial neural network (Ann) algorithm is used to replace the original retrieval algorithm in CBR system, and the BBN training classifier is used to establish the IDSS model of emergency information based on ANN-BBN-CBR. Finally, the workflow of the system is given. The validity and feasibility of the system are verified by the collected data. [results / conclusions] the use of the system model not only improves the retrieval and matching speed and accuracy of large case base, but also improves the accuracy of the system. It also provides appropriate statistical information for the measurement of case similarity, and also brings convenience to the organization and management of the case base, and provides research ideas for the collection, integration, analysis, transmission and feedback of emergency information.
【作者单位】：西北工业大学管理学院;西安电子科技大学经济与管理学院;
【基金】：国家自然科学基金项目“非常规突发事件下管理者情绪对应急决策的影响机制研究”(编号:71171162) 陕西省社会科学基金项目“陕西省重大工程项目社会稳定风险评估机制及防范策略研究”(编号:2015R004)
【分类号】：G350;TP18

【相似文献】