运用时序多指标决策的专家库动态优化  

Dynamic Optimization of Expert-Base Based on Time Series Multi-Attribute Decision Making

作　　者：李佳瑾[1] 郭鹏[1] 

机构地区：[1]西北工业大学管理学院,陕西西安710072

出　　处：《工业工程》2011年第2期57-60,69,共5页Industrial Engineering Journal

基　　金：国家自然科学基金资助项目(70672103)

摘　　要：针对同行评议在专家评审方面的局限性,指出了对专家工作业绩进行静态评价的缺陷,提出了对专家的工作业绩进行实时追踪的思想,应用时序多指标决策方法并在选定所需的"时间度"的基础上,从业绩指标的好坏程度和业绩指标的变化情况两个角度,对专家业绩进行动态评价,进而达到对专家库进行动态优化的目的。Peer review is a widely recognized approach.It is an expert review process and the performance is greatly dependent on the selection of experts that is done by using an expert-base.The information in an expert-base is often obtained by statically evaluating the experts＇ performance,but not dynamically updated.In this paper,the drawbacks are analyzed for such a review process.Based on the analysis,to solve this problem,we propose to dynamically update the expert-base by tracking their real-time performances.With time series multi-attribute decision making,dynamic experts＇ work performance evaluation is presented by choosing the appropriate time scale.Thus,the expert-base can be dynamically updated based on not only the static indexes but also their changes.In other words,the expert-base is dynamically optimized.

关 键 词：同行评议 业绩评价 时序多指标决策 专家库 

分 类 号：O212.2[理学—概率论与数理统计] C816[理学—数学]