See https://elliit.liu.se for a general description on ELLIIT.

In the ELLIIT subproject Quality assurance in continuous software engineering, we collaborate with LU/LTH, Axis and Ericsson to explore the interaction between automated and manual data analysis in continuous software engineering.

Advances in technology, like cloud computing and modern tool-chains for automated builds, testing and deployment, have enabled organizations engaging in continuous software engineering to deploy a new version of a system ever more rapidly. This new way of working requires automation and puts new requirements in terms of the role and responsibilities of quality assurance.

In this subproject, we will explore the interaction between automated and manual data analysis in such contexts. We will investigate the use of data analytics and visualizations to help software engineers interpret the massive amount of data available due to activities like code analysis, version management, code reviews, testing and product usage. Furthermore, to support the integration of research activities and results in the industrial context, we will advance the work on improving the knowledge co-creation between industry and academia in software quality assurance.

In the ELLIIT subproject Gazing at code reviews, we also collaborate with LU/LTH and Axis to study to investigate how intelligent assistance, driven by eye-tracking, may assist developers in reviewing code.

In software development, developers spend a lot of time reading and understanding code. One key comprehension activity in a software developer’s workflow is modern code review. Reviewing code takes a lot of time, but companies consider code reviews to be important enough to merit the costs since it not only is a tool for software quality assurance but also for knowledge sharing.

In this project, we investigate how we could help developers in making code reviews more efficient and more effective. Ideally, this means providing developers with information adapted to their current needs and helping them to focus on the task at hand. We, therefore, study eye movements during code review to detect elements (beacons) that trigger certain activities or steps in comprehension processes. The goal is to eventually build and evaluate prototypes that assist software developers in reviewing code.

See https://elliit.liu.se for a general description and links to subprojects at other sites.