Toward Requirements and Design Traceability Using Natural Language Processing

Abstract : The paper aimed to address the problem of incompleteness and inconsistency between requirements and design stages, and how to make efficient linking between these stages. Software requirements written in natural languages (NL), Natural Language Processing (NLP) can be used to process requirements. In our research we built a framework that can be used to generate design diagrams from requirements in semi-automatic way, and make traceability between requirements and design phases, and in contrast. Also framework shows how to manage traceability in different levels, and how to apply changes to different artifacts. Many traceability reports can be generated based on developed framework. After Appling this model we obtained good results. Based on our case study the model generate a class diagram depends on central rule engine, and traceability was built and can be managed in visualize manner. We proposed to continue this research as its very critical area by adding more Unified Modeling Language(UML) diagrams, and apply changes directly inside software requirement document.
Complete list of metadatas

https://hal.laas.fr/hal-02005396
Contributor : Abd-El-Kader Sahraoui <>
Submitted on : Monday, February 4, 2019 - 9:50:36 AM
Last modification on : Saturday, October 26, 2019 - 1:35:15 AM

Links full text

Identifiers

Citation

Abd-El-Kader Sahraoui, Omer Salih Dawood. Toward Requirements and Design Traceability Using Natural Language Processing. European Journal of engineering research and science, 2018, 7 (3), ⟨10.24018/ejers.2018.3.7.807 ⟩. ⟨hal-02005396⟩

Share

Metrics

Record views

41