# Artificial Intelligence helps making Quality Assurance processes leaner

**Authors:** Alexander Poth, Quirin Beck, Andreas Riel  
**arXiv:** [1906.02970](https://arxiv.org/abs/1906.02970)  
**Date:** 2019-06-07  
**Summary page:** https://www.summarizepaper.com/en/paper/1906.02970v1/

## Key points

- Lean processes and artificial intelligence (AI) are discussed in optimizing test management.
- Regression tests are important to ensure functionality from previous releases still works in new releases.
- Many regression tests are currently manual, making them expensive and time-consuming.
- The authors propose a semi-automated approach using machine learning (ML) as a support tool for test management.
- ML is used as a supporting machine while the human test manager selects appropriate test cases.
- The selection of regression tests is crucial and specific to each release, considering changes such as new code or features.
- Different approaches can be used to identify the right set of regression tests, including source code file level methods and test case prioritization techniques for black-box tests.
- Continuous prioritization trends in regression testing are important, and the authors aim to support it with their ML-based approach for black box regression test case prioritization.
- The paper presents an improvement in the test process through the use of AI and ML as support tools, aiming to make quality assurance processes leaner and more efficient while involving human expertise.

## Summary

This paper discusses the use of lean processes and artificial intelligence (AI) in optimizing test management. It introduces the concept of regression tests, which are performed to ensure that functionality from previous releases still works in new releases. However, many regression tests are not automated and are executed manually, making them expensive and time-consuming. To address this issue, the authors propose a semi-automated approach using machine learning (ML) as a support tool for test management. The proposed lean testing process aims to use ML as a supporting machine while keeping the human test manager responsible for selecting appropriate test cases. The selection of regression tests is crucial and specific to each release, requiring the consideration of changes such as new or deleted code, refactoring or implementation of new features. The task of deciding which tests can be skipped based on relevance is typically performed by the test manager with expert support. To identify the right set of regression tests different approaches can be used depending on the software process. Common entry points for identification include source code file level methods while advanced approaches combine multiple file level methods. For black-box tests techniques like test case prioritization can be employed. The authors highlight the importance of continuous prioritization trends in regression testing and aim to support it with their ML-based approach for black box regression test case prioritization. In summary, this paper presents an improvement in the test process through the use of AI and ML as support tools for test management. By semi-automating the selection of regression tests they aim to make quality assurance processes leaner and more efficient while still involving human expertise in decision-making.