基于信息融合的罪案关联证据推理模型

针对传统犯罪案件中出现的冲突证据难以处理的情况,提出一种基于证据可信度和灰色关联的冲突证据融合处理方法;基于相似性视角的灰色关联度作为证据间的联系,采用改变证据源的证据组合规则来衡量证据源中各个证据之间的贴近度;考虑到证据的可信度,提出新的基于证据贴近度的权重确定方法,以证据的可信度作为权重,对参与融合的证据的基本概率分配函数进行加权平均,使证据融合收敛速度更快,并提升融合效果.最后以安徽省某市的入室盗窃犯罪案件为例,运用基于信息融合的关联证据推理方法处理案件中的冲突证据问题,验证了所提出方法的合理性和有效性.

Crime-related evidence reasoning model based on information fusion

For the situations that it''s difficult to deal with the conflicting evidences in traditional criminal cases, a conflicting evidences fusion processing method based on evidence credibility and grey incidence is proposed. Based on the degree of grey incidence on similarity perspective as a link between evidences, a combination of evidences that changes the source of evidences is used to measure the closeness between the various evidence sources. Considering the credibility of evidences, a new weight determination method is proposed based on the closeness of evidences. Taking the credibility of the evidence as the weight, the weighted average of the basic probability assignment function of the evidence participating in the fusion is used to make the evidence fusion convergence faster and improve the fusion effect. Finally, through the case of a theft crime in a city of Anhui province, the evidence reasoning method based on information fusion is used to deal with the conflicting evidences in the case, which reflects the rationality and effectiveness of the proposed method.