Association Rule Mining for Requirement Elicitation Techniques in IT Projects

Abstract

Selecting suitable techniques for requirements elicitation in IT projects is crucial to the business analysis planning process. Typically, the determining factors are the preferences of stakeholders, primarily business analysts, previous experience, and company practices, as well as the availability of sources of information. The influence of other factors is not as evident. One of the possible ways to form recommendations for using techniques is the analysis of industrial experience. This paper is intended to analyze the application of association rules mining to define factors influencing technique selection and predict the usage of a particular elicitation technique depending on the project context and specialist background. The dataset for experiments was formed based on a survey of 328 specialists from Ukrainian IT companies. The associations found to make it possible to speed up the process of choosing elicitation techniques and improve the elicitation process efficiency.