Natural Language Processing for Requirements Engineering

Natural Language Processing for Requirements Engineering (NLP4RE) is an area of research and development that seeks to apply natural language processing (NLP) techniques, tools, resources the requirements engineering (RE) process, support human analysts carry out various linguistic analysis tasks on textual documents, such as detecting issues, identifying key domain concepts, establishing traceability links. This article reports a mapping study surveys landscape NLP4RE provide holistic understanding field. Following guidance systematic review, directed by five questions, cutting across aspects research, concerning state literature, empirical focus, tool development, usage NLP technologies. Our main results are follows: (i) we identify total 404 primary studies relevant NLP4RE, which were published over past 36 years from 170 different venues; (ii) most these (67.08%) solution proposals, assessed laboratory experiment or example application, while only small percentage (7%) in industrial settings; (iii) large proportion (42.70%) focus phase, with quality defect detection their central task specification commonly processed document type; (iv) 130 tools (i.e., RE specific tools) extracted studies, but 17 them (13.08%) available download; (v) 231 technologies also identified, comprising 140 66 25 resources, them—particularly those novel techniques specialized tools—are used infrequently; contrast, traditional (e.g., POS tagging tokenization), general-purpose Stanford CoreNLP GATE) generic lexicons (WordNet British National Corpus). The not provides collection literature also, more importantly, establishes structure frame existing through categorization, synthesis conceptualization theoretical concepts relationships encompass both aspects. work thus produces conceptual framework NLP4RE. gaps directions, highlight technology transfer needs, encourage synergies between community, one, software systems practitioners. can be starting point future according well-defined terminology expanded new solutions emerge.