Abstract
Effective software maintenance requires a detailed knowledge of the system's artifacts, the way these artifacts are used or modified and their interrelationships. Based on some useful characteristics of the object-oriented paradigm the identification of objects within procedural programs has become a promising approach to reduce the effort in program understanding and, hence, the maintenance cost. In this paper we present a new approach to object identification in procedural programs that not only relies on information exclusively extractable from source code but integrates human expertise and external domain- and application-specific knowledge.

This publication has 17 references indexed in Scilit: