| Summary: | 碩士 === 國立成功大學 === 資訊工程學系碩博士班 === 93 === Due to the fact that an ontology can describe the specific domain knowledge with the concepts and relationships and play an important role in many information systems and Semantic Web, therefore a great deal of the specific domain knowledge deriving from the documents in which knowledge represents and applies will be an important research topic. In this thesis, we adopt the characteristic of the ontology to describe the domain-specific knowledge of Chinese documents to achieve requirement tracking and Chinese document summarization tasks. With regard to the process of Chinese requirement document we propose a method, Ontology-based Requirements Traceability Agent, adopts Concept Transformation, Relation Generator, Relation Weight Generator, and Traceability Ontology Generator with Chinese requirement document to construct a Traceability Ontology. As requirements may change, the Requirements Traceability Agent can effectively analyze the impact of the changes according to the structure of traceability knowledge. With regard to the process of Chinese news document we propose a method, Fuzzy Ontology-based Chinese document Summarization, first derives term set of Chinese news document from Document Preprocessing Mechanism, and Term Classifier, and then utilizes the Fuzzy Inference Mechanism to transform a Domain Ontology into a Fuzzy Ontology. Finally, the Chinese Document Summarization Agent refers to Fuzzy Ontology to address Document Preprocessing Mechanism, Fuzzy Inference Mechanism, Genetic Learning Mechanism, Sentence Path Generator, Sentence Generator, and Sentence Filter to deal with document summarization. The experimental results show that our approach can effectively address impact analysis during the requirements of the changes and Chinese document summarization.
|
|---|
