多源评价中信息标准化研究

Study of information uniformization course in multi-source evaluation

在基于自然语言的群决策中,由于自然语言的不确定性,首先需要确定自然语言所属的集合。在这个集合中,通过语言值与其他语言值之间的差异,这个语言值的语义可以唯一地确定。而在一个评价模型中使用多个自然语言集是经常遇到的情况。在处理这类模型时,利用某个标准语言值重述语言值的语义是必需的。在重述过程中,如果不相似性度量保持一致,那么信息标准化过程就可以保持语言值的语义不变。在现有许多文献中,对于标准语言值集合的选取规则并未做过多的要求,只是设定此集合的势在7到13之间。为了进一步控制语言值语义在重述过程中的改变,提出了标准语言值适用性度量这一概念。

In group decision making model based on linguistic infolmation,because of the uncertainty of natural language,it is needed-to find the linguistic term sets which are used to express the information.In the sets, the meaning of linguistic term is uniquely confirmed through the difference from others.A series of nature linguistic term sets are used in an evalua- tion model,In the approach of aggregating the linguistic term in different sets,it is necessary to redescribe linguistic terms us- ing a group of standard linguistic terms,In the redescribing course,if the dissimilarity measurements are kept,the information uniformization course will hold on the meaning of linguistic terms.In many articles,the cardinality of the set of standard lin- guistic terms is set from 7 to 13,but other material rules which used to choose those important linguistic terms in evalua- tion system are not provided.In order to decrease the difference of linguistic term in the uniformation course, the conception of the applicability measurement of standard linguistic terms set is given.