Software requirements specification analysis using natural language processing technique

Understanding software requirements has been widely acknowledged as a crucial task in software development projects. This problem emerges since software requirements, which are originally specified using a particular natural language, are often ambiguous and incomplete. The condition will be completely different if the requirements are specified using formal language in which ambiguity and incompleteness could be obviously found and thus quickly anticipated. Transforming “natural” software requirements into a more formal specification may therefore reduce their ambiguity and incompleteness. This paper hence proposes a method to transform software requirements specified in a natural language to formal specification (in this context is object-oriented specification). The proposed method uses natural language processing technique.