Predictive Skill Based Call Routing Using Multi-Label Classification Techniques

Abstract

This article is in the context of a call centre whose agents possess a heterogeneous skillset. The significant challenge for such a call centre would be skills based call routing: to match an inbound customer call to a call centre agent possessing the relevant skillset. This article will present an alternative to the usual Interactive Voice Recording (IVR) menu based approach to skill based call routing. This article will also make use of a multi-label classification techniques to predict the purpose of the customer call in advance and route it to the appropriate call centre agent without the customer's intervention. This hassle-free call routing technique produces efficient interactions and helps enhance customer experience, resulting in higher customer satisfaction and better cross-sell opportunities.