Abstract: In real applications, the degree of membership of formal contexts may change over time. However, the existing dynamic concept cognition methods can only deal with the concept cognition problem when the objects and attributes of the formal context change incrementally. When the  degree of membership of the formal context changes, the existing dynamic concept cognition methods cannot achieve cognition. In order to remedy this deficiency, a dynamic concept cognition method is proposed when the threshold value of the cognitive operator and the partial  degree of membership of the formal background change. Specifically, for the case of threshold change, the incremental operator is used to transform the original concept into the updated concept of the formal background；for the case of  degree of membership change, the changed formal background is divided into three parts by using the different change cases, and the cognition is performed on the three subformal context respectively, and the updated concept of the formal background is finally generated. Then, a cognitive system that can handle the simultaneous changes of threshold and  degree of membership of the cognitive operator is established. Finally, the concept cognition algorithm is given, and the effectiveness of proposed algorithm is illustrated by data experiments.

Key words: dynamic conceptual cognition, cognitive operators, cognitive systems, fuzzy formal context, degree of membership

摘要： 在现实应用中，形式背景的隶属度可能随着时间发生变化。然而，现有的动态概念认知方法只能处理形式背景的对象和属性增量变化情况下的概念认知问题。当形式背景隶属度发生变化时，现有的动态概念认知方法不能实现认知。为了弥补这一不足，提出一种认知算子的阈值、形式背景的部分隶属度发生变化情况的动态概念认知方法。具体地，对于阈值发生变化的情况，利用增量算子将原概念转化为形式背景更新后的概念；对于隶属度发生变化的情况，利用变化情况的不同，将变化后的形式背景分成三个部分，分别在三个子形式背景上进行认知，最终生成形式背景更新后的概念。建立了一个可以处理认知算子阈值、隶属度同时发生变化的认知系统。给出了概念认知算法，并通过数据实验说明了算法的有效性。

关键词: 动态概念认知, 认知算子, 认知系统, 模糊形式背景, 隶属度