How Chatbots in Financial Services are Evolving with NLP and NLU

Published on June 11, 2018

Active.Ai Elizabeth Duke

SVP of Business Development, North America, Active.Ai

AI-powered chatbots can understand your specific banking needs using NLU — and also detect how you feel and predict what you’ll say next.

Processing human language is a challenge for artificial intelligence — the way we converse is nonlinear, irregular, emotional, and full of context. The goal is for AI to hold a back-and-forth conversation with a human in a way that feels natural — despite the fact that it’s with a machine. Natural language processing — or NLP — is the first step to making this a reality.

In customer service-focused industries, like banking, chatbots equipped with NLP can analyze, process, and communicate with users, using language they understand. NLP techniques categorize customer data by tagging parts of speech, correcting spelling, and re-formatting numbers and dates into something the machine can read.

Even so, NLP by itself can’t address the complexities of human communication. Our conversations are non-linear, and it’s not enough for a chatbot to lead us through a rigid step-by-step process. Customer expectations have moved beyond what most web and mobile interfaces can deliver. Not every user has the same preferences or talks in the same way — in the real world, our conversation is unstructured with jargon and short-form, and we go off on tangents or repeat ourselves.

To better understand these complexities, chatbots need natural language understanding — or NLU. This is how Active.Ai takes the conversational experience to a deeper level.


How NLU augments NLP: the difference made by natural language understanding

NLU is an advanced subset of NLP that enables a machine to understand and respond to complex conversation formats. For a chatbot to communicate effectively with customers, it needs to be able to handle the ebbs and flows of real conversation. With advanced NLU, a chatbot can keep the conversation going by using clarification techniques (“Did you mean this or that?”), providing facts, detecting sentiment, or predicting what the customer will ask next.

A major benefit of NLU is the ability to understand context. Think of a scenario where you’re accessing your bank account and you say, “Yes, that will balance things out.” What if the only word picked up on was “balance,” and you were then directed to your account balance? This wouldn’t be helpful, and could cause frustration on the part of the user. The subtleties of context are a challenge for AI-powered chatbots, but NLU provides a deeper level of understanding.


NLU solves traditional customer pain points

You’ve most likely had a frustrating experience with interactive voice response (IVR), where a menu-driven voice makes you fill in the blank in order to receive an answer or direct your call. It’s only a matter of time before you hit “0” to try and talk to a human being.

Chatbots that use NLP without NLU will come up against similar pain points — the conversation doesn’t feel natural. Much of what is seen in the market is simple slot-filling where the bot draws out information from the user, step-by-step. With NLU, the onus is taken off the customer and supported by the machine instead. The user can speak however they normally would — regardless of structure or regional expressions — and the chatbot’s job is to understand and take appropriate action.

Specifically for banking, at Active.Ai we bring a deep understanding of financial services that is embedded into pre-trained user journeys and datasets — this gets a customer to market faster with the right solution, whether it’s applying for a credit card or securing a loan. It eliminates customer frustration, and doesn’t turn them off of using a chatbot as a customer service channel in the future.


AI-powered chatbots need NLU to succeed

Chatbots have the power to provide 24/7 personalized customer service and meet the user wherever they are. By applying NLU capabilities, it’s easier to detect customer preferences, opinions, feelings, and inclinations, and learn how to enhance the quality of interactions while building insights to deepen customer relationships.

At Active.Ai, we understand what financial institutions need. Our Triniti AI engine is easy to deploy and equipped with advanced NLP and NLU to interpret intent, sentiment, and emotions of your customers for personalized conversation. Request a demo today.


Source: Medium

Contact Us

Request a Demo