结合可信度模糊评价与动态聚类的构件选择

摘要

Internetware is faced with the significant challenges of trust in the open and dynamic Internet environment.A component selection approach meeting trust expectation nearest to user requirement is explored by means of the fuzzy theory.In this approach six trust attributes of component are defined for Internetware,and a fuzzy synthetic method of multi-factors is applied to evaluate component's trust.On the basis of these,an Internetware model to satisfy the trust evaluation expectation of user requirements is built,and a trusted component selection method employing dynamic clustering based on fuzzy equivalence relation is explored to achieve the mapping from candidate component to abstract component.An application case is given to show the effectiveness of this approach.%开放、动态的Internet环境下,网构软件面临可信性的重大挑战.运用模糊理论,提出了一种满足最贴近用户可信度期望的构件选择方法.该方法中,定义了网构软件环境下构件的6种可信属性,介绍了一种多因素的构件可信度模糊综合评价方法,以此为基础,建立了一种满足用户可信度期望的网构软件模型,为实现候选构件与抽象构件的映射,应用基于模糊等价关系的动态聚类实现可信构件的选择.结合案例说明了方法的有效性.