SOFTWARE TESTING USING MACHINE LEARNING

摘要

A system and method for analyzing (104) a computer program includes performing a static analysis on a program to determine property correctness. Test cases are generated (108) and conducted to provide test output data. Hypotheses (113) about aspects of execution of the program are produced to classify paths for test cases to determine whether the test cases have been encountered or otherwise. In accordance with the hypothesis, new test cases are generated (120) to cause the program to exercise behavior which is outside of the encountered test cases.