On XML Based Automated Function Point Analysis: An Effective Method to Assess Developer Productivity

Abstract

This article proposes a novel approach to objectively measure developer productivity and individual contribution to development projects using automated function point analysis (integrated with source control system (SCS) and continuous integration (CI)). Automated function point analysis is a central tenant required in this approach. The method proposed herein relies on pattern matching to identify and group fields in to internal logical files, external inputs, external outputs, and external query objects which can be used to identify the objective size of software. Unlike earlier automated functional sizing algorithms, the method proposed is applicable to any application source code that expresses its user interface in well-formed extensible markup language (XML). This study shows that a prototype implementation of the proposed automated counting method is capable of providing a reliable and consistent relative size of existing software (accurate to +/-6.65% and 5000 times faster), and goes on to explain how this method can be integrated with existing requirements engineering tools, source control systems and continuous integration tooling in order to produce an objective measure of software developer output compared to estimates established during requirements gathering and planning that exceeds the common, subjective means that are in use in many organizations today.