A Non-Exclusive Multi-Class Convolutional Neural Network for the Classification of Functional Requirements in AUTOSAR Software Requirement Specification Text

Abstract

Software Requirement Specification (SRS) describes a software system to be developed that captures the functional, non-functional, and technical aspects of the stakeholder's requirements. Retrieval and extraction of software information from SRS are essential to the development of software product line (SPL). Albeit Natural Language Processing (NLP) techniques, such as information retrieval and standard machine learning, have been advocated in the recent past as a semi-automatic means of optimising requirements specifications, they have not been widely embraced. The complexity in the organization's information makes requirement analysis intricately a challenging task. The interdependence of subsystems and within an organisation drives this complexity. A plain multi-class classification framework may not address this issue. Hence, this paper propounds an automated non-exclusive approach for classification of functional requirements from SRS, using a deep learning framework. Specifically, Word2Vec and FastText word embeddings are utilised for document representation for training a convolutional neural network (CNN). The study was carried out by the compilation of manually categorised relevant enterprise data (AUTomotive Open System ARchitecture (AUTOSAR)), which were also employed for model training. Over a convolutional neural network, the impact of data trained with Word2Vec and FastText word embeddings from SRS documentation were compared to pre-trained word embeddings models, available online.

Automatic derivation of use case diagrams from interrelated natural language requirements

“…On the other hand, word embeddings can be obtained by using the totality of the development data as training input. It is worth mentioning, that Chatterjee et al (2020) and Jp et al (2022) report improved results in the subsequent grouping step with the implementation of self-trained embedding models. On the contrary, Gulle et al (2020) achieve superior results with pre-trained embeddings compared to self-trained.…”

A Systematic Literature Review on Using Natural Language Processing in Software Requirements Engineering

“…The identification of actors and actions through NLP and artificial neural networks (ANNs), with the aim of enhancing precision in requirement analysis, were studied by Cascini et al (2004) [96], Imam et al (2021) [47], and Al-Hroob et al (2018) [51]. Convolutional neural networks (CNNs) were applied for functional requirements classification within SRS documents [36], for advancing categorization granularity [82,87] or even for SRS document classification [50]. Deep learning methods were also employed in SRS-quality-assessment activities [85] to automate these activities using K-means, NLP, case-based reasoning, and multi-agent systems [33,87], or to predict and understand SRS quality attributes [86,108].…”