Abstract: The design of SDML is on the basis of spatialSQL in spatial database, extending operators in spatial data mining while using language of data mining process model for reference. SDML can be divided into view manipulation language(VML) and model manipulation language (MML) by different steps of data mining process and objects that SDML manipulates. VML is responsible for manipulating views of data mining, while MML is for models of data mining. The thoughts and design of SDML are described in detail, and a solution is presented that use SDML for spatial generalization and spatial association, which are two typical problems in spatial data mining.

Key words: spatial data mining, data mining language, data mining view, data mining model

摘要： 设计了一种基于空间数据库的空间数据挖掘语言SDML。根据SDML操作的对象以及挖掘过程的不同阶段，SDML语言可以分为视图操纵语言和模型操纵语言，分别负责对于数据挖掘视图和模型的操作。详细阐述了SDML的设计思想及其设计方案，针对空间泛化和空间关联这两个典型的空间数据挖掘问题，给出了SDML解决方案。

关键词: 空间数据挖掘, 数据挖掘语言, 数据挖掘视图, 数据挖掘模型