Design and use of neural network applications in telecommunications

Abstract

This chapter describes the use of neural networks in the analysis of software systems. The development of large software systems over long periods of time, and the software crisis that this has produced, provides a rich source of problems. Many of these problems may be addressed, in part at least, with the aid of pattern matching and pattern classification tools such as artificial neural networks. We concentrate on two problems: analyzing the complexity of large software systems following the procedure set out in Reference 1 and detecting copies (clones) of source code procedures in large software systems.