Measuring Similarity of Large Software Systems Based on Source Code Correspondence

摘要

It is an important and intriguing issue to know the quantitative similarity of large software systems. In this paper, a similarity metric between two sets of source code files based on the correspondence of overall source code lines is proposed. A Software similarity MeAsurement Tool SMAT was developed and applied to various versions of an operating system(BSD UNIX). The resulting similarity valuations clearly revealed the evolutionary history characteristics of the BSD UNIX Operating System.