To say that human conversations are complex is an understatement. Our communication prowess depends on where we learnt a language, our teachers, our educational level, and other factors. More complexity arises when conversations involves persuasion, negotiation, and other threads involving a back and forth exchange – that must be tracked.
Thankfully, common knowledge possessed by humans is a critical prop to understanding a conversation. For example, a customer quickly grasps a contact center agent’s pitch to offer a telecom plan at a competitive price. At this point the customer may negotiate and receive a counter offer. Both parties are aware of negotiation as a process. Here is a sample of a real conversational exchange.
If machines could similarly and accurately understand contextual detail it would amplify productivity at contact centers. For example, a supervisor of customer service agents could instantly obtain a granular analyses of sub optimal product pitches – automatically for all conversations – and help her team members improve performance.
Machines have struggled to understand such conversations because current language engines typically depend on keyword-type searches for identification of language patterns. However, the emergence of knowledge-based A.I. and its application to Natural Language Understanding (NLU) has made it possible to accurately understand a complete, complex conversation.
AUI Systems is applying this knowledge-based NLU accurately to conversations to identify risk, improve customer experience, and increase revenue at contact centers. We’ll delve deeper in Part 2 of this article.