Identification, Assessment and Automated Classification of Requirements Engineering Techniques

Abstract

Selection of suitable techniques to be used in requirements engineering or business analysis activities is not easy, especially considering the large number of new proposals that emerged in recent years. This paper provides a summary of techniques recommended by major sources recognized by the industry. A universal attribute structure for the description of techniques is proposed and used to describe 33 techniques most frequently quoted by reviewed sources. A pilot study of automated classification of techniques based on attribute values is also reported. The study used fuzzy c-means clustering algorithm and produced pairings of complementary techniques, most of which successfully passed validation conducted by business analysis practitioners.