Abstract: Rough ontology is a basic tool to handle imprecise information. Its form is a tree structured by the rough concept and the rough relation. The specific form of rough ontology presents the multiplicity because the different domain experts participate in constructing it, but the structure of rough concept lattice also composed of the rough concept and the rough relation is unique. The rough ontology construction can use similarity of the composition and the structure of lattice and tree. Construct a rough concept lattice; transform it to a rough ontology. This paper discusses the steps of a lattice-tree converse method: construct a rough concept lattice; cluster rough concepts in the rough concept lattice to generate a rough concept tree; transform the rough concept tree for a rough ontology. A constructing example illustrates the concrete application of the lattice-tree conversion method.

Key words: rough ontology, rough concept, rough concept lattice, rough concept tree

摘要： 粗糙本体是处理不精确性信息的一种基础性工具，其存在形式是由粗糙概念和粗糙关系构成的树形结构。粗糙本体的具体形式因为参与构建的领域专家的不同而呈现多样性，然而同样也是由粗糙概念和粗糙关系构成的粗糙概念格却具有结构上的惟一性。粗糙本体的构建因而可以利用格和树的组成与结构的相似性，构建粗糙概念格，转化生成粗糙本体。探讨了格-树转化方法的实现步骤：构建粗糙概念格；聚类粗糙概念格中的粗糙概念生成粗糙概念树；转化粗糙概念树为粗糙本体。构建实例阐释了格-树转换方法的具体应用。

关键词: 粗糙本体, 粗糙概念, 粗糙概念格, 粗糙概念树