Text mining is an effective means of detecting potentially useful knowledge from large text documents. However conventional text mining technology cannot achieve high accuracy, because it cannot effectively make use of the semantic information of the text. Ontology provides theoretical basis and technical support for semantic information representation and organization. This paper improves the traditional text mining technology which cannot understand the text semantics. The author discusses the text mining methods based on domain ontology, and sets up domain ontology and database at first, then introduces the “concept-concept” correlation matrix and identifies the relationships of conceptions, and puts forward the text mining model based on domain ontology at last. Based on the semantic text mining model, the depth and accuracy of text mining is improved.