基于改进二元语义的评估信息融合研究  

Improved 2-Tuple Linguistic Model Based Data Fusing Method for Evaluation

作　　者：侯彩虹[1] 丁永生[2] 

机构地区：[1]上海海关学院管理系,上海201204 [2]东华大学旭日工商管理学院,上海200051

出　　处：《系统仿真学报》2011年第10期2142-2146,共5页Journal of System Simulation

基　　金：上海市高校选拔培养优秀青年教师科研专项基金(shg08001);国家哲学社会科学资助项目(11CJY024)

摘　　要：针对评估中评估结果表达形式多样不便比较分析,以及个体评估结果分散与最终评估结果统一间矛盾的存在,在此提出基于二元语义模型的评估信息融合方法,该方法利用改进的二元语义模型对数据标准化,在此基础上引入融合算子,将表达形式不同的多评估结果有效融合,最后给出评估小组的群决策结果。用该方法对海关风险评估问题进行仿真,结果表明该方法可以有效地完成评估问题中从数据预处理、数据分析到结论支持整个过程。专家数据融合的实例验证了所提出模型的可行性和有效性。In the evaluation, there are some existing problems, for example, different forms of evaluation results are not easy to be compared and analyzed, and information from different evaluators need to be fused. With respect to above problems, an improved 2-tuple linguistic model based aggregation method for solving evaluation information was proposed. With the 2-tuple linguistic model based aggregation method, the main aimed at using aggregation operator to fuse all information to satisfy the needs of group-decision in evaluation. In the model, the whole process was implemented, i.e., pre-solve basic, and data and draw conclusion was analyzed. The practical experiments of data processing for customs risk evaluation demonstrate the flexibility of the model.

关 键 词：二元语义 数据融合 融合算子 风险评估 

分 类 号：TP934[自动化与计算机技术]