Incorporating Domain Knowledge into Pattern-based Requirements Engineering for Improved Quality Assurance and System Verification

Conceptos Básicos

This approach integrates domain knowledge into pattern-based requirements engineering to enable automated quality assurance and support system verification through defect-based testing.

Resumen

The paper proposes an Approach to Pattern-based Domain-Specific Requirements Engineering (DSRs) that incorporates domain knowledge into the requirements specification process. The key elements are:

DSRs use graphical representations (figures and tables) to capture domain-specific terminology and parameters, in contrast to traditional textual requirements.
The meta-data associated with DSR elements enables semi-automated quality checks based on domain-specific defects, addressing issues like omission, imprecision, and contradiction.
The DSR specification is integrated with defect-based testing techniques for system compliance verification, particularly for performance requirements with continuous input spaces.

The authors demonstrate the approach using the example of take-off performance requirements for Unmanned Aerial Vehicle (UAV) controllers. They show how the domain knowledge can be represented in the DSR elements, enabling quality assurance checks and supporting the defect-based testing process.
The proposed approach aims to close existing gaps in requirements engineering by incorporating domain knowledge, improving requirements quality, and integrating requirements engineering with system verification activities.

Estadísticas

The take-off performance DSR specifies the following parameters:

Initial pitch (Θ) angle: 0 to 20 degrees
Initial yaw (Ψ) angle: 320 to 360 degrees
Initial roll (Φ) angle: 0.001 to 0 degrees
Initial height (h): -1 to 2 meters
Maximum horizontal deviation (Δ(x,y)): 2 meters
Maximum altitude overshoot (Δhov): Specified at different time steps after reaching the target height

Citas

"The incorporation of domain-specific knowledge in DSRs is valuable in defining precisely these defects. For the Take-Off Performance DSR, it is defective if one specifies unrealistic parameters without rationale."
"The greater the defect-value, the greater is the chance for the scenario Si that generated Ξ to trigger a defect for the actual UAV."

Ideas clave extraídas de

Towards an Approach to Pattern-based Domain-Specific Requirements Engineering

by T.Ch... a las arxiv.org 04-29-2024

https://arxiv.org/pdf/2404.17338.pdf

Consultas más profundas

How can the proposed approach be extended to handle non-functional requirements, such as safety and security, in addition to performance requirements?

To extend the proposed approach to handle non-functional requirements like safety and security, the Domain-Specific Requirements (DSRs) framework can be adapted to incorporate specific patterns and elements related to these aspects. For safety requirements, DSRs can include elements that define critical safety parameters, constraints, and failure modes. Security requirements can be integrated by introducing DSR elements that specify authentication mechanisms, data encryption standards, and access control policies. By defining domain-specific patterns for safety and security within the DSR framework, the approach can ensure that these non-functional requirements are captured effectively alongside performance requirements.

What empirical studies or industrial case studies could be conducted to validate the effectiveness of the approach in practice?

Empirical studies can be conducted in collaboration with industry partners working on projects involving complex systems development. These studies can involve applying the Pattern-based Domain-Specific Requirements Engineering Approach to real-world projects and evaluating its impact on requirements quality, system verification, and overall project success. Industrial case studies can focus on domains like automotive systems, medical devices, or financial technology, where stringent requirements and domain-specific knowledge play a crucial role. By comparing the outcomes of projects using the proposed approach with traditional methods, the effectiveness and benefits of the approach can be empirically validated.

How scalable is the approach, and how can it be applied to other domains beyond UAV controllers, such as self-driving cars or other AI-based systems?

The scalability of the approach lies in its flexibility to adapt to different domains by defining domain-specific patterns and elements within the DSR framework. To apply the approach to domains beyond UAV controllers, such as self-driving cars or AI-based systems, domain experts from those fields can collaborate to identify relevant requirements patterns and incorporate them into the DSRs. By customizing the DSRs to reflect the unique characteristics and requirements of each domain, the approach can be seamlessly extended to address the specific challenges and complexities of diverse domains. This adaptability ensures that the approach can be effectively applied to a wide range of industries and systems beyond UAV controllers.