A knowledge and data base for software systems

Abstract

We discuss a knowledgddata centered architecture for CASE environments, that has at its heart a knowledge and data base management system (repository) on which all CASE tools rely for knowledge and data services. The repository has three entity kinds: objects, procedures, and rules.Data resides in the objects while knowledge resides in the procedures, rules, and special fields in the objects. We then discuss why a third generation DBMS, which incorporates data, object, and knowledge services, is a good choice for the knowledge manager.We use an object oriented model of software systems to analyze the storage requirements of program documents in the repository, and show how we use the procedures and rules to efficiently maintain multiple software document representations. We then show how the repository can effectively support a software reuse library, as well as reuse processes throughout the life-cycle. We conclude by outlining the future directions of our research.

Using Relational Databases in the Engineering Repository Systems

“…Lack of powerful type system. Taylor et al (1988), Miguel et al (1990), Emmerich et al (1992), Liu et al (1996), Gray (1997). Poor support to data that represents graph structures.…”

“…Lack of facilities for making queries based on such data including finding transitive closure. Hardwick (1984), Miguel et al (1990), Katz (1990), Emmerich et al (1992), Gray (1997), Lange et al (2001).…”

“…Linton (1984), Miguel et al (1990), Chen et al (1990), Lange et al (2001). Inappropriate transaction models.…”