Abstract: Concerning the categorical characteristics of spatial data, a rule-based spatial co-location method in categorical data is introduced. This algorithm defines the transaction of data mining by using spatial relation and finds the co-location rules by using the multi-layer participation index. It can solve the problem effectively. Some experiments have been made on fire data in a city by this method. The range of applications and the performance of this approach are analyzed, and it proves that the rule-based spatial co-location approach in categorical data is applicable to other continuous data after being discreted, with good results obtained.