A Tool for Detecting Ambiguity in Software Requirements Specification

Abstract

The main goal of requirements engineering is to establish software requirements specification (SRS). The requirements in SRS are mostly specified in natural languages (NL), therefore, one of the common problems of SRS is requirements ambiguity. The requirement is said as ambiguous when it has more than one interpretation, subsequently, can lead to requirements inconsistency and conflict. Besides, to detect ambiguous requirements manually is time-consuming and tedious process. Thus, this paper presents a tool called SRS Ambiguity Detector, that able to detect automatically the major types of ambiguity; lexical, syntactic and syntax ambiguity. This tool uses ambiguity words from the ambiguity handbook to detect lexical ambiguity, while, parts of speech (POS) tagging technique has been applied to detect syntactic and syntax ambiguities. Evaluation was conducted to assess the effectiveness, and the result has shown that the proposed tool able to identify more ambiguous requirements as compared to manual detection.