Software Functional Requirements Classification Using Ensemble Learning

Vijayvargiya, Sanidhya; Kumar, Lov; Malapati, Aruna; Murthy, Lalita Bhanu; Misra, Sanjay

doi:10.1007/978-3-031-10548-7_49

Cited by 1 publication

(1 citation statement)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The lack of interpretability in some ML-based approaches poses challenges in understanding and validating the reasoning behind classification or prediction outcomes (Habib et al, 2021;Rahman et al, 2019). Lastly, the limited real-world implementation of NLP and ML approaches in industry settings indicates a gap between research and practical application (Cheligeer et al, 2022;Vijayvargiya et al, 2022a).…”

Section: What Are the Limitations Presented In The Existing Research?mentioning

confidence: 99%

A Systematic Review on Natural Language Processing and Machine Learning Approaches to Improve Requirements Specification in Software Requirements Engineering

Embi,

Why

2023

ijmst

View full text Add to dashboard Cite

This systematic literature review (SLR) examines the current practices, challenges, proposed solutions, and limitations of natural language processing (NLP) and machine learning (ML) approaches in improving requirements specification in software requirements engineering. The review focuses on research conducted in the last five years and includes a selection of papers that discuss the use of NLP and ML techniques for enhancing the accuracy and clarity of requirements, particularly in the context of functional and non-functional requirements. The findings highlight the benefits and challenges associated with the integration of NLP and ML approaches, such as improved classification and identification of requirements. However, it is observed that there is a greater emphasis on non-functional requirements, with a limited representation of research on functional requirements. Comparison of this review and the recent two reviews has been done to observe the differences and highlight the novelty and contribution. The review also identifies limitations, potential bias in assuming that problems related to requirements documentation or specification can be easily resolved through simple changes as well as the need to address the functional requirements. The insights from this SLR contribute to the understanding of the current state of research in this field and provide a foundation for future research directions and practical applications in leveraging NLP and ML approaches to enhance requirements specification in software requirements engineering.

show abstract