模糊WINGS视角下的ANP加权矩阵新构造方法

构造因素集加权矩阵是ANP系统未加权超矩阵到加权超矩阵转化的一个关键技术。然而从已有的三种构造方法看,两两比较法比较机理混乱,而等权矩阵假设构造法通常是无效的,并且基于DEMATEL（决策试行与评价实验室）的构造方法不仅存在忽视因素集自身强度的内在不足,而且也难以反映专家在对因素集之间影响关系判断时存在的“不精确性”。为克服上述缺陷,提出一种模糊WINGS（加权影响情景下的非线性测度体系）视角下的ANP因素集加权矩阵新构造方法。该方法一方面给出改进后的模糊DELPHI决策程序,充分考虑了专家判断过程中的“不精确性”。另一方面,系统提出模糊WINGS的方法思路,在系统因素集影响关系判别时充分考虑了因素集的自我影响强度,使因素集直接影响矩阵的构造更为合理。实例对比验证结果表明,该方法是科学合理的,有着较强的实践应用可操作性。     

基于区间直觉模糊集的Sinarchy超矩阵构造方法改进

针对网络分析法（ANP）中Sinarchy结构的超矩阵构造问题,提出了一种基于区间直觉模糊集的超矩阵构造改进方法。从Sinarchy结构中具有反馈关系的最后两层元素集的固有属性及特征出发,分析了采用传统方法构造方案对准则的影响矩阵（IMAC）的局限性,提出了应结合整体论与还原论、并遵循钱学森倡导的从定性到定量的综合集成法作为ANP分析解决问题的思路。通过引入区间直觉模糊集代替传统点估计值反映专家偏好进而重新构造IMAC。具体实例应用表明,该方法针对Sinarchy结构的排序问题具有较好的实际应用可行性。     

Evaluating Componentized Enterprise Information Technologies: A Multiattribute Modeling Approach

Corporations are increasingly implementing enterprise information technologies (EITs), because of the costs of maintaining legacy systems and the lack of fit of such systems for organization-wide information sharing. A new type of EIT that is being introduced in major corporations (such as Dell Computers), is based on the idea of component systems, which are stand-alone software programs that can integrate with other such components with relative ease. Given the financial outlay for EITs, the evaluation and adoption of these systems is not something that can be completed haphazardly. This requirement is complicated by the relative infancy of models for the evaluation of componentized EITs. To this end, in this paper, we introduce a managerial multistage multiattribute decision model, consisting of the Analytical Hierarchy Process (AHP) model and the Supermatrix approach (also defined as the Analytical Network Processes (ANP) approach). This combination of models builds on earlier work that validates the Supermatrix approach for evaluating traditional EITs at a Fortune 100 organization. The aggregation of these benefits is then measured against the costs of systems, thereby arriving at a ranking of alternatives for each functional area. We illustrate the model with an example and draw managerial implications.