Once you have confirmed all requirements in the document you will like to analyze, and have ensured that QVscribe has been properly configured (by either pressing Configuration in the QVscribe toolbar, or Customize Your Analysis under the Analysis button), press Analyze # Requirements at the bottom of the main QVscribe window, (‘#’ will actually be the number of confirmed requirements in the document). Note that QVscribe currently has a limit of 3000 requirements for analysis.
Once you press on Analyze # Requirements a pop-up window will appear showing the analysis in progress. The duration of the analysis will depend on the number and size of requirements confirmed. If the analysis appears to take a long time (many minutes) or appears to not be progressing, cancel and try the analysis again with fewer requirements.
QVscribe performs three distinct types of analysis on the selected requirements, each focusing on a key area:
1. Requirements Quality: Each requirement is assessed and scored based on established requirements engineering standards and best practices using the following set of Key Quality Indicators:
- Negative Imperatives
These quality indicators are assessed based on the presence of user-configurable trigger words defined in the QVscribe Configuration.
2. Term and Unit Consistency: For term consistency analysis, QVscribe detects and highlights all noun-phrases found in the requirements and displays them along with the requirements containing them, and along with any similar terms. This helps verify that domain-specific terminology is consistent throughout the document and that the correct terms are used in each requirement. For unit consistency analysis, QVscribe detects and highlights all units in the selected requirements based on the list of units in the QVscribe Configuration. The detected units are displayed by type, along with the requirements containing them. This helps assess that all units are correct and consistent throughout the requirements.
3. Requirements Similarity: QVscribe analyzes the syntax of each requirement and shows any requirements found that is similar to it. The level of similarity can be tuned higher or lower to help assess if duplicates or contradicting requirements are present.