Literature Review on Lexical Ambiguity Detection Techniques in Software Requirements Specifications

Abstract

A number of studies note that the cumulative percentage of a software development process failure is due to failure to anticipate the requirements specification. One such failure is the ambiguity in the Software Requirements Specification (SRS) document. For that the process of fulfilling the needs of software to be one of the most important processes in the manufacture of software. Detecting ambiguity from software requirements spesifiaction manually is a way that tends to generate many errors, takes time, and costs a lot. There are many techniques and approaches that can be used to detect ambiguity, of course, these techniques have different paths with each other. In this study, the authors raised the topic of lexical ambiguity detection technique in the Software Requirements Specifications (SRS) document. Researchers reviewed journals from previous research to refer to a conclusion about the techniques used in detecting lexical ambiguity. In this study refers to several points of view, namely the number of stages and steps used in each technique, factors that play a role in the implementation of the technique and also the level of accuracy of the results of techniques used. The results of this study suggest that the Natural Language Processing technique is the most accurate technique used in detecting lexical ambiguity in SRS documents in fulfilling software requirements.