Research and Development (R&D) Project Management has been recognized as an essential knowledge management task for modern organizations and society. With the development of web-based R&D project management systems, funding agencies now face a daunting challenge to manage huge volume of electronic R&D project data available online. Meanwhile, ontology has been widely studied in recent knowledge management and information system research. Inspired by the success of ontology applications to the reuse of knowledge embedded in scientific documents in scientific community, this research proposes an ontology-based approach to R&D project management. To create a domain specific ontology, the following three issues are critical, (1) semiautomatically creating a domain specific ontology, where advanced techniques in ontology engineering and soft computing can be exploited; (2) maintaining an unambiguous specification of concepts or relations; and (3) establishing inference mechanisms to allow knowledge reasoning. These issues will be addressed in this thesis. First, a bi-lingual fuzzy ontology development framework is proposed to automatically create domain specific ontologies using various techniques drawn from research areas of ontology engineering, information extraction and fuzzy logic. Second, a novel hybrid approach to domain ontology construction is developed. Concepts and relations are automatically leant from training R&D project documents. Third, the inference mechanism of fuzzy ontologies is investigated and a novel Fuzzy Conceptual Matching method is developed for fuzzy expert ontology construction. Finally, the proposed framework and methods are implemented and applied in a prototype system for potential use in Mainland China. The major contributions of this thesis are: (1) a novel bi-lingual fuzzy ontology development framework which provides a unified and extendable theoretical framework to underpin the research into ontology-based R&D project management; (2) a novel domain specific bi-lingual fuzzy ontology which provides a formal and semantically rich representation of domain knowledge and information; and (3) a novel fuzzy inference engine which is defined and developed for expert ontology construction, bridging between document space, concept space and expert space.