Motivating Natural Language Processing Part 1: Innovative Chatbots for Serious Businesses

Matthew Roberts

Motivating Natural Language Processing Part 1: Innovative Chatbots for Serious Businesses

Do you have a great customer experience? With so many ways to reach your customers, this answer may be hard to pin down. Even so, most companies are now differentiating themselves on the basis of customer experience (CX), and innovative companies are finding new and creative ways to compete in this ever-changing landscape.

Natural language processing (NLP) provides a set of powerful tools for those looking to compete on CX. Offering flexibility, scalability, creativity, and personalization, NLP has the potential to make a dramatic impact on your business.

Still, you may not have a clear understanding of NLP or how it can be applied to improve your customer experience. We’re kicking off this series as a guide through the NLP landscape and to help innovative companies add these powerful, exciting tools to their toolbelt.

What Is Natural Language Processing?

Natural language processing is so much more than chatbots—though don’t worry, we will talk about chatbots.

More than anything else, natural language is data. It’s noisy and chaotic; it’s structured (though in deeply complex and fascinating ways) and often open to interpretation; but still, it’s data. It can be cleaned; it can be analyzed and visualized; and it can drive insights and actions for those who know how to interpret it.

NLP is the attempt to make sense of all this data. Long confined to the halls of academia, it has reached an inflection point and is ready to explode onto the scene. With the help of a robust and ever-expanding NLP toolset, innovative companies can activate this powerful data set to carry their businesses into the future. Before we get started, here are a few key terms you should know.

Key Terminology

Speech-to-text – The process of converting speech input into digital text, based on speech recognition. Speech-to-text (S2T) is often the first requirement for a successful NLP project utilizing human voice.
Document summarization – The process of shortening a set of text computationally to present a subset with key, relevant information that is representative of the whole.
Document tagging – The process of categorizing a set of text computationally with key words, topics, or other identifiers, often to enable grouping or discoverability among sets of documents.
Intent recognition – The process of assigning some meaning or intent to a voice or text utterance. This is a key component in most chatbots to enable complex responses to a variety of user inputs.
Named-entity recognition – A kind of information extraction process that assigns words or phrases to certain predefined categories to make sense of raw text data.
Sentiment analysis – The processes of assigning a positive or negative score to a set of text indicating the writer’s attitude toward or perception of the object of discussion.

We will touch on each of these topics in more detail throughout the course of this series, highlighting their capabilities and demonstrating their technical implementations. For now, let’s talk about chatbots.

Chatbots: More Popular and Getting Smarter

You could be forgiven for thinking that chatbots might be entirely synonymous with NLP. Chatbots were one of the first and still are the most widely adopted use case for NLP. Though they have grown more robust and mature over the past few years, many of the fundamental technologies and processes we use when designing for conversation remain the same.

What has changed is user adoption. A 2018 report by PWC shows that 90% of consumers are familiar with voice assistants, with 72% having used one themselves. Even so, in many cases poor conversational design and a lack of trust have prevented all but the most basic of use cases from gaining much traction.

These challenges are solvable. With inventive thinking and a concern for the consumer experience, innovative companies can harness the power of the conversational user interface (UI) to transform their user experience and engage their customers in new and meaningful ways.

Innovating With Conversational User Experience

Just as a credit card form that looks stuck in the 1990s can leave customers worried about the security of an ecommerce site, a clunky chat experience can damage consumer trust for companies looking to engage their users in this new medium.

On the other hand, putting the same care and concern into your conversational user experience as you do in the rest of your user interface can transform your customer experience and serve as a key differentiator for your business. We’re going to walk through some specific ways to leverage NLP for an enhanced user experience.

Design Thinking

As with many products, you often will not know how effective your chatbot is until you put it in the hands (or mouths) of your users. Bad assumptions made in the design process can lead to a broken user experience.

This is the perfect use case for design thinking. Rapid ideation and prototyping, followed by user acceptance testing, can tighten the feedback loop on your chatbot design and shine a light on potential pitfalls in your user experience.

Integrated Audio-Visual Interface

Some information is simply not suitable for voice. Anyone who’s had to sit through a lengthy call center menu can attest to this. On the flip side, the richness of natural language opens the doors to a world of possibilities for user input and navigation within visual interfaces trapped in the realm of point and click.

A fully integrated audio-visual interface enables both mediums to play to their strengths and allows users to engage in a way that’s convenient for them.

NLP and IoT

The rise of embedded devices and the internet of things (IoT) has allowed powerful technology to become as mobile as we are. Any home with smart speakers or other smart devices has seen the benefit this technology can have on our daily lives. NLP is the perfect companion for this new, interconnected world.

Natural language serves as an intuitive and infinitely customizable interface for these IoT devices. Pre-built language models, tuned to your users and environment, can be used offline to support your interaction with these devices, without tying them down to a finicky (or nonexistent) internet connection.

Read more of our thoughts on designing effective chatbots here.

Chatbots: More Than a Flashy Gimmick

Chatbot projects are often dismissed as marketing ploys—all flash with little to no substance. While chatbots are certainly fun to build and more fun to interact with, they are serious business, and innovative companies have already taken note.

According to Gartner, 25% of companies have already invested in chatbots to support their customer service operations. This makes sense. One Walker report noted that 91% of companies that are effective at customer experience have a competitive advantage. NLP is one of many tools used to take advantage of this customer experience opportunity.

Speed and Personalization in Customer Service

Chatbots allow your customer service process to scale to meet demands, reducing the amount of time your customers will spend in frustrating waiting rooms. Full automation is not always the answer though. Instead, real human agents, freed from supporting trivial customer requests, can now serve as a backstop for an automated process to service your customers’ most challenging needs.

Supporting Health and Safety for All Users

We’ve already seen the impact of COVID-19 on many disruptive technologies. NLP is perfectly positioned to support a “contactless” yet personalized interface with your customers. From doctor’s office check-ins to restaurant ordering, conversational user interfaces support the health of all users by reducing human-to-human touch points that could spread disease.

More than just health, conversational UI can be employed to address a number of other safety concerns. Manufacturing and automotive industries, where “eyes on the road” can be a matter of life or death, can use NLP to interact with workers and drivers without the visual distractions that screens might provide.

Accessible and Engaging Products

ADA compliance requires many businesses to adapt the way they interact with their users. Chatbots provide an excellent alternative to a visual medium to support the perceivability and operability of your products.

More than compliance, an engaging conversational UI can dramatically improve the customer experience for users who are unable to interact with your products visually.

Get Help With Your Chatbot

How do you know if a well-conversed bot is right for your business? Assessing your users and their pain points is a great start. We’d love to help! Get in touch with us at findoutmore@credera.com to learn more.