Abstract
The validation and verification of software is typically a costly part of the development. A possibility to reduce costs is to concentrate these activities on the fault-prone components of the system. A classification approach is proposed that identifies these components based on detailed UML models. For this mainly existing code metrics are tailored to be applicable to models and are combined to a suite. Two industrial case studies confirm the ability of the approach to identify fault-prone components.
| Originalsprache | Englisch |
|---|---|
| Seiten (von - bis) | 435-452 |
| Seitenumfang | 18 |
| Fachzeitschrift | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
| Jahrgang | 3463 |
| DOIs | |
| Publikationsstatus | Veröffentlicht - 2005 |
| Veranstaltung | 5th European Dependable Computing Conference, EDCC-5 - Budapest, Ungarn Dauer: 20 Apr. 2005 → 22 Apr. 2005 |