A software process data model for knowledge engineering in information systems

摘要

Knowledge engineering for information systems is a long-term, multi-person task that requires tight control and memorization not only of what knowledge is acquired but also of why and how it is acquired. We propose a software process data model as the foundation of a knowledge-based software information system that emphasizes control, support and documentation of design decision-making and tool integration in information systems environments.The model is developed along two dimensions. Firstly, it defines how to represent and integrate design objects (what), design decisions (why) and design tools (how). Secondly, it exploits the abstraction mechanisms of the extensible hybrid knowledge representation language CML/Telos to manage the evolution not only of particular software projects, but also of the software development environment in which these projects operate. Modular aggregation relates design-in-the-small and design-in-the-large support. Besides motivating and formalizing the model, we describe an operational prototype implementation called ConceptBase and report intitial application experiences in the DAIDA ESPRIT project.