Identification of propagated defects to reduce software testing cost via mutation testing.

Area of Science:

• Software Engineering
• Software Testing
• Software Maintenance

Background:

• Software testing is a costly but crucial part of software maintenance.
• A single test round may not detect all software defects, increasing maintenance costs.
• Mutation testing is used to assess test suite effectiveness.

Purpose of the Study:

• To propose a new approach using mutation testing to identify defects missed by single-round testing.
• To reduce software testing costs by improving defect detection efficiency.
• To enhance the effectiveness of test suites in software maintenance.

Main Methods:

• The proposed method applies multiple mutants to a program simultaneously.
• It analyzes relationships between code lines based on mutant behavior.
• Defects are identified by examining these relationships against single-round testing results.

Main Results:

• The new method was evaluated against seven common fault localization techniques.
• A comparative study used the widely-used Defects4J dataset.
• The proposed approach demonstrated superior performance compared to existing methods.

Conclusions:

• The proposed mutation testing approach effectively identifies defects missed in single testing rounds.
• This method offers a promising way to reduce software testing costs.
• It outperforms established fault localization techniques, enhancing defect prediction accuracy.