NLP Tool for the Detection of Ambiguities in Software Requirements Written in Spanish

EasyChair Preprint 9671, version 2

Abstract

Requirements engineering is one of the most important stages of the software development life cycle. The success of any software product depends on the quality of its requirements. Software requirements are usually written in natural language. Ambiguity in requirements written in natural language is a problem that has been studied by the requirements engineering community for more than two decades. Manual resolution of ambiguity in requirements is tedious and time consuming. Several natural language processing tools exist to automate ambiguity analysis, however, most of them are not widely available, obsolete, unsafe, and expensive; the few public tools only support English language requirements analysis. This research aims to develop a natural language processing tool to detect lexical and syntactic ambiguity present in software requirements, using the Python programming language and natural language processing tools such as NLTK. As a result of this work, a dataset containing 19,357 requirements belonging to software development projects at the University of Computer Science is presented; the obtained dataset constitutes a baseline for future research. The XP methodology was used to guide the development of the proposed tool. The approach was evaluated on a data set of 100 requirements and we achieved 98% accuracy and 91% completeness

Keyphrases: Ambigüedad, Procesamiento del Lenguaje Natural, Requisitos de software, Técnicas, dataset

@booklet{EasyChair:9671,
  author    = {Samira Enriquez and Jonathan Ramírez Reyes and Dunia Colomé Cedeño and Reiman Alfonso Azcuy and Héctor González Diez},
  title     = {NLP Tool for the Detection of Ambiguities in Software Requirements Written in Spanish},
  howpublished = {EasyChair Preprint 9671},
  year      = {EasyChair, 2023}}