基于AHP数据处理的科研绩效评价模型  

Scientific Research Performance Evaluating Model Based on the Treatment of AHP's Data

作　　者：余宏杰[1] 焦立新[1] 窦鹏[1] 

机构地区：[1]安徽科技学院,安徽凤阳233100

出　　处：《安徽科技学院学报》2008年第4期28-33,共6页Journal of Anhui Science and Technology University

基　　金：安徽省教育厅自然科学研究项目(2006KJ054C)

摘　　要：多指标评价广泛存在于日常决策活动中,它是评价人员对多个目标方案进行综合衡量后做出合理的、正确的评价过程。各指标权系数的估价及赋值,相当程度上决定了多指标评价的精度。层次分析法(简称AHP),是一种将半定性、半定量的问题转化为定量计算的方法,是一种定性与定量相结合的多目标决策分析方法,它特别适用于那些难于完全用定量进行分析且缺乏必要数据的复杂问题。针对传统的层次分析法的不足,利用"波动系数平均值"来确定控制线,筛选数据、剔除异常值,提高权重系数的准确率。由此得出科研绩效评价一种新模型。The multi -indicator evaluation is widely used in the daily decision -making, and it is the evaluation process through which the. personnel could appraise reasonably and correctly after they balance many goal plans comprehensively. To a considerable extent, the accuracy of multi - indicators evaluation was determined by the estimate and the evaluation for each indicator coefficient. The analytic hierarchy process （ i. e. AHP） is a method which transforms the semi-qualitative and semi -quantitative question as the quantitative evaluation, but al- so it is a multi -objective decision -making analysis method which unifies with the both. It is particularly suitable for those complex problem which is difficultly analysised by completely using the quantitative analysis method, and the absence of essential data. Being the shortcoming of the traditional AHP, using the mean of fluctuant coefficient, the authors set a controlling limit. The AHP＇s data were strictly selected, and the exceptional value were removed, so that the accuracy of weight was improved. Thus the model for evaluating the scientific research performance was obtained.

关 键 词：层次分析法 波动系数 科研绩效 评价模型 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]