Knowledge representation support for a software information system

Peter G. Selfridge · 2002

CODE-BASE is a software information system that uses knowledge representation (KR) technology to represent a wide spectrum of knowledge about a body of telecommunications software and uses several techniques to guarantee the knowledge base is synchronized with the code. In addition, CODE-BASE allows a user to extend the underlying KR to capture discovered relationships between the code and the user's domain model. The author describes the classic knowledge representation system used to build CODE-BASE. He then describes CODE-BASE itself, its representation of knowledge and its interface, and briefly illustrates its capabilities. He then discusses in detail how the KR system supports the CODE-BASE functionality, including the representation of diverse kinds of knowledge and the support for evolution, as well as several other advantages.>

Read the paper · More papers on PaperTik