Geo 101: What Is Natural Language Processing?
Here's what marketers need to know about the field of computer science making it possible for voice-activated assistants to understand spoken requests — and then answer them.
With the goal of breaking down some of the most important concepts to provide a better understanding of the basics — and a jumping off point for exploring how far technology may take us — we introduce the next installment of our GeoMarketing 101 series: what marketers need to know about Natural Language Processing (NLP).
What Is Natural Language Processing?
Natural language processing is a branch of artificial intelligence related to how software can ‘listen to,’ process, and manipulate language. Put simply, it’s all about is the ability of a computer program to understand human language in the way that it is spoken.
This might sound a bit futuristic, but just about everyone who has picked up an iPhone is actually familiar with NLP: It’s what makes it possible for voice-activated assistants like Siri or Alexa to understand spoken requests and then answer them. (Natural language processing does refer to the interpretation of both speech and text — anything written or said to a computer that isn’t in a programming language — but voice is the present disrupter and new frontier for NLP.)
Current approaches to NLP are “based on deep learning, a type of AI that examines and uses patterns in data to improve a program’s understanding” of language over time, Margaret Rouse writes for Search Engine Analytics. “Deep learning models require massive amounts of labelled data to train on and identify relevant correlations, and assembling this kind of big data set is one of the main hurdles to NLP currently.”
It’s a significant challenge. After all, it was “just recently — in fact, just less than two years ago — [that] voice recognition passed the threshold that makes it a viable user interface,” Dan Quigley, STO, senior manager for Alexa Smart Home, told GeoMarketing last March. “That’s all [about] accuracy.”
Essentially, as natural language processing improves, so too does the ability for seamless communication and understanding between people and their devices — and the removal of this friction opens up new possibilities for consumers and marketers alike.
Why Does NLP Matter To Marketers?
Whether they’ve been thinking about it or not, NLP affects marketers’ business: first, because of the ability of voice-based communication to engender relationships, and second, because of the growing import of understanding the intent behind users’ queries when it comes to search marketing.
The “rise of voice search” has perhaps seen more than its fair share of press this year. But its growth — and marketers’ approach to it — is critical first and foremost for a very simple reason: Voice-based interaction has been proven to be cognitively simpler for humans than text/swipe, making it feel intuitive and more personal to users. Its NLP that makes this kind of communication between a person and their device — and ultimately, a brand, a consumer, and a device — possible.
Additionally, and perhaps most importantly, is the reason behind why more brands may ultimately opt to explore NLP themselves: While understanding queries has always mattered to search marketers, the new world of search — in which customers look for structured answers to their queries in featured snippets or from a voice assistant, rather than clicking on a series of blue links — prizes understanding the intent behind a users’ query. This means that marketers need to get smarter about intent, and AI is a compelling way to do this.
After all, people don’t always say what they mean clearly — but today, a brand still needs to 1. provide a structured answer and 2. understand what kind of related inquiries a person might make so that they can determine what content might be related to the original search.
Yext’s Duane Forrester put it like this: “Voice search is very much about ‘ask a question, get an answer.’ It’s not just about, “Let’s go do some keyword research and focus on keywords now.” This has expanded into the concept of topics: If I am purchasing a new Xbox, what else might I be purchasing? I might be purchasing HDMI cable. I might be purchasing an extra controller. I may want games. I’m probably going to want a subscription online. So there are at least four other discreet elements there that are directly related to the purchase someone is making that, as a marketer, I need to be talking about [in my content.]
“It’s very important to understand that there are discreet questions, and you need to have answers for them.”
Read more about voice, marketing, and NLP: