Task Models as a Mean to Identify and Justify Automations in Development Tasks

Multiple empirical studies have been proposed to support developers' tasks by means of automation. These automations exhibit different levels of granularity ranging from low-level tasks to much higher-level tasks. However, proposed automation focusses more on the functional aspects of the development tools rather than the user aspect (e.g. goals and work) leaving the supported tasks under specified. We propose to use and extend the HAMSTERS task-modeling notation to describe developers' goals and tasks. We demonstrate on simple examples how the task models can be used to identify good candidates for automation and to assess the benefits and drawbacks brought by automation.