• Requirements Management Flashcards
(11 cards)
Requirements Management: Core Tasks
▪ Rationale Management and Traceability:
* Justification of requirements
* Linking requirements to each other and to development artifacts
▪ Requirements Management:
* Structuring, documentation and archiving
* Attribution
▪ Interactions with other management tasks:
* Validation and verification
* Change management including impact analysis
* Version management
* Configuration Management
* Claim Management
* Support for distributed RE
Elementary Attributes: Priority and Acceptance Criterion
Priority:
▪ Not always all requirements can be implemented and usually not all requirements are equally
important
→ Prioritization of requirements
Acceptance Criterion:
▪ It is not always possible to clearly evaluate and check all requirements, or to decide during
acceptance when the implementation of the requirement was appropriate from the perspective of
the stakeholders
→ Acceptance criterion defines criteria under which the implementation of the requirements is
considered valid
Procedure for handling corrections regarding requirements according to their specification:
Apply -> Rate -> Decide -> Realize
Objectives of change management for requirements
▪ Systematic and controlled recording of change requests
▪ Evaluation of change requests (sense and effort, risks etc.)
▪ Decision on the implementation of the changes
▪ Implementation of the changes
Establishment of Change Management
Define a change process
Create responsibilities and a change control board
Creation of a change database that supports the change process
what are Quality Gates
Quality Gates are quality measurement points
Constructive Quality Assurance
Elicitation Techniques, Specification Techniques
(Standards, Processes), Prototyping
Analytical Quality Assurance
Dynamic Techniques, Static Techniques
Smells in Requirements
Smell detection is used for manual and automatic quality assurance
▪ Problem: Issues in requirements, such as ambiguities or incomplete requirements specifications,
can lead to time and cost over-run
Why NLP? Common errors in specifications
Textual specifications are fundamental for SW development processes
Yet (remember user stories as an invite to converse)
* Often bad quality due to using natural language
* Problems: Inherent ambiguities, inconsistencies, etc.
Quality model for specifications
atomicity
unambiguity
conciseness
testability
traceability
consistency
formal correctness