Research Article
Recovering Software Design from Interviews Using the NFR Approach: An Experience Report
Table 5
Validation results.
| Item | Positives | Scope for improvement |
| Component and connector view | 90% correct | No central broker; Information Brokering Service (IBS) is more important than SBS; bidirectionality assumptions are incorrect. |
| Activity diagrams | 60% activities correct; 88% links correct | Some of the links had wrong sources and/or destinations; some of the activities were in the wrong order. |
| Logical deployment model | 100% correct | This has sparse information. |
| Physical deployment model | 90% correct | We concluded SpringSource was used for application server—in fact, Java Services Container does this job. |
| Detailed views of services | 93% correct | Forwarder should be Output Channel Manager. |
|
|