Deriving relevant functional measures for automated development projects

Abstract

The increasing use of computer aided software engineering (CASE) tools, fourth-generation languages (4GLs) and other similar development automation techniques, has reduced the impact of implementation methods and individual ability on development task difficulty. It has therefore been suggested that measures derived from software specification representations may provide a consistent basis for relatively accurate estimation of subsequent development attributes. To this end, this paper describes the development of a functional complexity analysis scheme that is applicable to system specification products, rather than to the traditional products of the lower-level design and construction phases.