A Proposal for Simplified Model-Based Cost Estimation Models

摘要

Most cost estimation models require a measure of the functional size of the application to be developed. To this end, FPA (Function Point Analysis) is one of the most used functional size measurement methods. FPA was originally proposed for traditional data processing systems, but it has been successfully adapted also to measure real-time and embedded systems. Since functional size measurement according to FPA can be quite expensive and time consuming, researchers have proposed "simplified" processes, which are expected to provide reasonably accurate measures, but require less effort and time. In this paper, we illustrate the application of these simplified techniques to UML models of software, via a precise mapping between UML elements and the so-called Basic Functional Components, upon which FPA measurement is based. As a result, it is possible to decrease the cost of modeling, and consequently the cost of measurement and estimation. The relatively low cost of the estimation models also allows developers to build different alternative models, to perform what-if analyses and choose the most economically sensible option.