针对在经典信念理论框架下,信念收缩后可能出现信息损失的问题，本文提出了一种利用遗忘理论来构建收缩算子的信念收缩方法。本文先通过理论证明来说明该收缩算子能够满足AGM理论中信念收缩的假定，然后用实例说明，与命题逻辑表示的信念遗忘收缩相比，一阶谓词逻辑表示的信念遗忘收缩保留了更多的原有信息，避免了不必要的信息损失，遵循最小修改原则。

A method of belief contraction based on the forgetting theory is presented for the information loss after belief contraction under the framework of the classic belief theory. At first, this method is proved theoretically to be able to meet the assumption of the AGM theory. Then through examples, it is illustrated that the belief contraction based on the forgetting theory expressed by the firstorder predicate can save more old information and obey the minimal change principle.