对科研项目进行合理排序,是科研项目评审和筛选的前提,也是科研管理的重要内容。被评科研项目数量有限情形下采用专家排序方式更为合理,而有序投票模型是处理此类问题的经典方法。提出了基于公共权重的有序投票模型,给出了基于交叉评价与总体评价的公共权重确定方法,弥补了传统有序投票模型中权重不一致、区分度不高的缺陷。
Ranking is the base of research projects evaluation and selection,and also important to science and technology research management.In the cases that the number of research projects under evaluation is limited,expert ranked voting is more reasonable to be used,and RVM (Ranked Voting Model) is a classic technique to deal with these problems.The paper proposes CWRVM (common-weights ranked voting model) and provides cross evaluation strategy and overall evaluation strategy to determine the common weights.The proposed methods can eliminate the limitations of various weights and low discrimination which exist in the traditional RVM model.