Automatically recommending test suite from historical data based on randomized evolutionary techniques

摘要

Disclosed herein are a system and a method for automated test suite optimization and recommendation, based on historical data, using randomized evolutionary techniques. The system analyzes historical data pertaining to file change pattern and test case execution history to identify test cases that match application being tested. Further, based on the test cases identified, the system generates optimized test suite recommendations to the user.

主权项

1. A method of automated test suite recommendation for an application, said method comprising:
collecting at least one input related to said application, using an automated test suite recommendation system; identifying at least one test case based on a historical data, wherein identifying said at least one test case using said automated test suite recommendation system further comprises:
identifying a file change pattern, wherein said file change pattern indicates a way files have changed across different builds of said application, using said automated test suite recommendation system;identifying test execution history corresponding to said identified file change pattern, wherein said test execution history indicates test cases changed according to said file changes across different builds of said application, using said automated test suite recommendation system;generating test case correlation value for each test case, based on said file change pattern and said test execution history, using said automated test suite recommendation system;selecting at least one test case based on said generated test case correlation value;deriving a regression coefficient based on said file change pattern and said test case execution history, using said automated test suite recommendation system; andcomputing test case recommendations based on said selected at least one test case and said regression coefficient, using said automated test suite recommendation system; and providing test suite recommendation to a user based on said at least one test case, using said automated test suite recommendation system.