Abstract | ||
---|---|---|
Many software development projects maintain repositories managing work items such as bug reports or tasks. In open-source projects, these repositories are accessible for end-users or clients, allowing them to enter new work items. These artifacts have to be further triaged. The most important step is the initial assignment of a work item to a responsible developer. As a consequence, a number of approaches exist to semi-automatically assign bug reports, e.g. using methods from machine learning. We compare different approaches to assign new work items to developers mining textual content as well as structural information. Furthermore we propose a novel model-based approach, which also considers relations from work items to the system specification for the assignment. The approaches are applied to different types of work items, including bug reports and tasks. To evaluate our approaches we mine the model repository of three different projects. We also included history data to determine how well they work in different states. |
Year | Venue | Keywords |
---|---|---|
2010 | ENASE 2010: PROCEEDINGS OF THE INTERNATIONAL CONFERENCE ON EVALUATION OF NOVEL APPROACHES TO SOFTWARE ENGINEERING | Machine Learning,Task Assignment,Bug Report,UNICASE,Unified Model,UJP |
DocType | Citations | PageRank |
Conference | 0 | 0.34 |
References | Authors | |
0 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jonas Helming | 1 | 123 | 9.03 |
Holger Arndt | 2 | 2 | 1.41 |
Zardosht Hodaie | 3 | 7 | 2.65 |
Maximilian Koegel | 4 | 142 | 10.22 |
Nitesh Narayan | 5 | 45 | 3.66 |