An ontology framework for evolution of plain text

Abstract

Data evolution is the process of semantics change for data product, in which data product has intrinsic semantics as well as extension semantics. This paper presents an ontology framework for evolution of plain text. The semantics of plain text is modeled as DATAGENOME ontology, including CONNOTATION ontology for expressing the intrinsic semantics and CONTEXT ontology for representing the extension semantics. CONNOTATION ontology is constructed according to the chapter structure of plain text. CHAPTER, SECTION, PARAGRAPH and SENTENCE class are used to describe a chapter, section, paragraph and sentence of plain text separately. CONNOTATION ontology may accompany with many CONTEXT ontologies. CONTEXT ontology consists of eight interconnected dimensions including what, when, where, who, how, which, why and awarings, is the description of virtual environment that lead to the existence of connotation. Instances of the CONNOTATION ontology can be viewed as a tree according to the relationship among them, so do the instances of COTEXT ontology. Mapping between connotation ontology tree and context tree show the correlation of intrinsic semantics and extension semantics.