中文摘要：

定性影响图是具有精确概率和效用的影响图的定性抽象。在定性影响图中，节点之间的影响关系使用定性符号描述，这种符号描述简化了不确定知识的表示，降低了影响图建模的难度，加速了不确定知识的推理。但是，定性影响图在抽象过程中损失了部分信息，导致定性影响图在评价过程中会产生不确定结果，阻碍了定性影响图的广泛应用。以加权的思想扩展定性影响图，使扩展的定性影响图中每个定性影响都带有一个表示节点问影响强弱的数值权值，在评价过程中根据定性影响权值的比较来减少不确定结果的产生，从而扩大定性影响图的应用范围。

英文摘要：

Qualitative influence diagrams （QIDs） are the qualitative abstraction of influence diagrams （IDs） with precise probabilities and utilities. In a QID, the influences between nodes are expressed by the qualitative symbols which simplify uncertain knowledge representation, ease influence graphs modeling and accelerate uncertain knowledge reasoning. However, during the abstraction process, qualitative influence graphs lose partial information, resulting in ambiguous evaluation results which impede the extensive application of QIDs. QIDs are extended by attaching a numerical weight to each qualitative influence of QIDs so that the ambiguous evaluation results can be avoided by comparing corresponding weights. Therefore, the application of QIDs will be broadened.