An effective sequential pattern mining algorithm to support automatic process classification in contact center back office

Contact center and its back office play a pivotal role on delivering excellent services to customer. However, back office process and operations become more and more complex, variable and costly due to frequent environment varying and the trend of staff-intensive. Automatic process classification and delimitation in back office is an effective way to help resolve these challenges, but it suffers very high deployment cost due to the complex and burdensome configuration works. In this paper, we propose an effective algorithm on sequential pattern mining to generate process patterns automatically, instead of manual configuration works, to achieve the goals of scalable deployment with high efficiency and low cost on automatic process classification and delimitation in contact center back office.