Skip to main content
Other
Categorization of Programs Using Neural Networks
Proceedings of the 1996 IEEE Symposium and Workshop on Engineering of Computer-Based Systems: Friedrichshafen
  • Franz J. Kurfess, New Jersey Institute of Technology
  • Lonnie R. Welch, New Jersey Institute of Technology
Publication Date
3-11-1996
Abstract

This paper describes some experiments based on the use of neural networks for assistance in the quality assessment of programs, especially in connection with the reengineering of legacy systems. We use Kohonen networks, or self-organizing maps, for the categorization of programs: programs with similar features are grouped together in a two-dimensional neighbourhood, whereas dissimilar programs are located far apart. Backpropagation networks are used for generalization purposes: based on a set of example programs whose relevant aspects have already been assessed, we would like to obtain an extrapolation of these assessments to new programs. The basis for these investigation is an intermediate representation of programs in the form of various dependency graphs, capturing the essentials of the programs. Previously, a set of metrics has been developed to perform an assessment of programs on the basis of this intermediate representation. It is not always clear, however, which parameters of the intermediate representation are relevant for a particular metric. The categorization and generalization capabilities of neural networks are employed to improve or verify the selection of parameters, and might even initiate the development of additional metrics

Disciplines
Publisher statement
Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.
Citation Information
Franz J. Kurfess and Lonnie R. Welch. "Categorization of Programs Using Neural Networks" Proceedings of the 1996 IEEE Symposium and Workshop on Engineering of Computer-Based Systems: Friedrichshafen (1996) p. 420 - 426
Available at: http://works.bepress.com/fkurfess/4/