TOWARDS ASSESSING THE QUALITY OF FUNCTIONAL REQUIREMENTS USING ENGLISH/SPANISH CONTROLLED LANGUAGES AND CONTEXT FREE GRAMMAR

Abstract

Most software applications are derived from customer requirements, which mainly are expressed as functional requirements. These requirements are the basis for software construction; therefore they have an important impact on overall software quality. As software grows so does the requirements, the increase in information often leads to overlooked requirement evaluation due to the amount of effort and time required. One important cause of this complexity in requirements evaluation is the fact that until now, most of them are still written in natural language which inherent all its problems to the requirements. In this paper we extended our previously proposed software requirement model which is based on context free grammar to support both English and Spanish under a controlled context for automated requirement evaluation using natural language processing, besides we provide comparative testing for both languages as well as comparison results for manual and automatic approaches. Keywords-Natural language processing; software engineering; software quality; software requirement