Software quality assessment based on semantic similarities

摘要

A method for software processing includes accepting quality information, which includes names of elements of software code and respective quality indications regarding tested acceptability of the elements. The names are processed to extract a list of substrings that occur in the names. Respective metrics are assigned to the substrings responsively to the quality indications of the elements in whose names the substrings occur. At least some of the substrings are presented to a user in accordance with the assigned metrics.