社交网络群决策环境下基于量子干涉的信任测度

针对已有的信任测度模型中信任源出现的冗余以及之间的干扰未被考虑的缺点,提出一种基于干扰效应的社会网络信任测度的决策模型.该模型在计算决策者之间的信任值时,通过简化信任源类型可避免部分信任源的重复出现,同时增加了信任源之间的干扰项.首先,根据社会关系与背景对信任的决定性,将决策者之间信任关系的来源划分为“决策者之间的亲近关系”和“决策者的客观背景”;其次,提出“亲疏度”和“专业度”的概念作为这两个信任源的测度,在保证信任源完整的前提下,克服信任关系冗余的缺陷;然后,进一步设计考虑干扰效应的信任值量化方法,所求的信任值能够综合体现两个信任源的个体决定程度和整体影响程度;最后,将决策者的信任值转化为权重聚集决策者偏好,以解决实际群决策问题.数值和理论结果表明,所提出考虑干扰效应的信任测度方法能够作为提高信任关系的准确性以及解决社交网络群决策问题的工具,具有广泛的应用前景.

Trust measure based on quantum interference in social network group decision making environment

In view of the redundancy of trust sources and the fact that the interference between them is not considered in the existing trust measure models, a decision model based on trust measure considering interference effect in a social network is proposed. When calculating the trust value between decision-makers, this model simplifies the types of trust sources, avoids the overlapping of some trust sources, and increases the interference item between trust sources. Firstly, the sources of trust between decision-makers are divided into ``close relationships between decision-makers'' and ``objective backgrounds of decision-makers'' according to the determination of social relationship and background on trust. Secondly, the concepts of ``closeness degree'' and ``professionalism'' are proposed as the measures of these two trust sources to ensure that the defect of redundancy can be overcome on the premise of integrity of trust sources. Then, a trust value quantification method considering the interference effect is further designed. The trust value can comprehensively reflect the individual decision degree and the overall influence degree of trust sources. Finally, the trust values of decision-makers are transformed into weights to aggregate the preferences of decision-makers to solve the actual group decision-making problem. Numerical and theoretical results show that the proposed trust measure method considering the interference effect can be used as a tool to improve the accuracy of trust relations and solve the group decision-making problems in social networks, which has a broad application prospect.