Software, Computer Applications
Observing, analyzing and understanding human factors is becoming a major concern in software development process in order to gain higher customer satisfaction. In this paper, we present a semi-automated methodology to generate the situation-transition structure which can be used to analyze the human behavior patterns in a specific domain. The term situation is defined as a 3-tuple < d; A;E > where d denotes human desire (mental state), A denotes the human actions vector, and E denotes the surrounding environment context vector. The situation-transition structure is a directed weighted graph where each node represents a unique situation or set of concurrent situations and an edge represents the transition from one situation to another. Data mining and machine learning techniques are used to generate situation-transition structure from raw observational data. We illustrate the proposed methodology through some case studies with open access datasets. The applications and advantages of situation-transition structure in software development are then asserted.