A method for improving bot effectiveness by recognising implicit customer intent in contact centre conversations

Contact centre systems are increasingly using intelligent voicebots and chatbots. These solutions are constantly evolving and improving. One of the main tasks of a virtual assistant is to recognise customers’ intentions. The accuracy of intent recognition should reach the highest possible level, because higher effectiveness of intent recognition translates into higher effectiveness of customer service. The purpose of this paper is to propose a new method for recognising customer intent. The functionality of this method takes into account emotions that accompany particular conversations, which is a new approach to the problem of recognising intent and determining the target flow of a conversation by bots. The method developed by the authors is intended for direct application in the customer service industry. The present paper analyses currently known solutions in terms of the intent recognition methods used. A linguistic analysis of actual conversations from contact centre hotlines was conducted for both the voice channel and the text channel. This analysis shows the clear impact of emotions that accompany conversations on the correctness of operation of virtual assistants. The results presented further in the paper show that it is possible to improve the correctness of the actions taken by bots by as much as above 50%. The solution presented herein is expected to improve customer service and optimise many KPIs.