Seizing Requirements Engineering Issues through Machine learning: A Systematic Mapping Study

Authors

  • María Guadalupe Gramajo Centro de Investigación y Desarrollo en Ingeniería en Sistemas de Información

Keywords:

Requirements Engineering, Software Requirements, Machine Learning, Supervised Learning

Abstract

The popularity of Machine Learning (ML) has grown exponentially. In recent years, researchers have exploited it to solve tasks and issues in Requirements Engineering (RE). This present paper shows a systematic mapping study, to provide a comprehensive review of researches detailing the application of ML techniques in RE to enrich traditional techniques and procedures. The results got highlight thirty-one (31) encouraging proposals aimed mainly at detecting ambiguity in requirements documents and their automatic classification into functional and non-functional requirements. The models most frequently mentioned by the analyzed proposals were Naïve Bayes, Support Vector Machine and Decision Tree.  We detected studies that still require further empirical validation to be accepted and applied by the software development community. This state of affairs denotes the need to research on applications, method, and performance of these techniques in RE.

Downloads

Download data is not yet available.

Published

2020-05-15

How to Cite

Gramajo, M. G. (2020). Seizing Requirements Engineering Issues through Machine learning: A Systematic Mapping Study. IEEE Latin America Transactions, 18(7), 1164–1184. Retrieved from https://latamt.ieeer9.org/index.php/transactions/article/view/54