Software testing is the key quality assurance mechanism in commercial software development and accounts for a substantial portion of the development costs. The increasing size and complexity of systems, shorter development cycles, and continuous deployment require effective and frequent testing. However, it becomes more and more difficult to make sense of the large volume of testing results.

The ability to interpret testing results is imperative for several decisions including (D1) deciding where to invest additional testing effort (i.e., test scoping) for cost-effective testing and (D2) assessing the confidence in the quality of a system for release (i.e., release readiness).

In practice, these decisions are based on heuristics (e.g. for decision D1: “test cases that have failed recently are likely to fail again”, “test cases that traverse the changed code are expected to reveal faults”, similarly for D2: e.g. “only X open issues of severity level Y”, “stable test outcome in the last Z builds”).

These decisions require practitioners to consider current and past data from additional internal and external sources (e.g., version control system for change impact analysis, defect reporting system, and customer usage statistics and feedback) to interpret the testing results within the broader software development context.

However, the amount of data and the number of sources to consider simultaneously makes it challenging for practitioners to process and interpret the results.

Project VITS uses data analytics and visualizations to support the interpretation of testing results and facilitates analytical reasoning to support the two key decisions D1 and D2.

For companies, the project will deliver support for two of the critical decisions in software development, which have a direct impact on their competitiveness in the market. From the scientific point of view, the project investigates the use of visualizations for supporting the decisions of test scoping and release readiness in practice.