NLP or Natural Language Processing is an area of science focusing on how people and technology interact. Yet what does this mean to contact centres?
The ability to talk to a computer instead of using a keyboard opens many cost saving opportunities as well as new and interesting interaction capabilities.
Yet for some industries it poses a new threat which will radically change how things are done.
As robots and automation revolutionise manufacturing and production, NLP looks set to revolutionise contact.
- Some background
- Using the technology
- What can you do today?
[read more=”Read more” less=”Read less”]
Some background
Talking to yourself
So what is the big deal, why even bother?
The reason is not everyone likes to type. Not everyone has the ability nor the patience for typing, so what if you could talk to your computer?
Insert Star Trek communicators and suddenly there’s a whole area of interest.
We know that mobile phones made it very easy to have portable technology.
They revolutionsied communication. They had challenges in the early days, now very few people on the planet don’t have a mobile phone. Those devices are getting smarter.
When driving it’s very difficult to text or look things up, say for instance a number for home. Even by having shortcuts it’s still frustrating if driving and you’re trying to get a number.
What if you could ask your phone to ring a number?
Already done and available technology wise, but the mass spark was lit with the advent of portable technology and a need to talk to devices.
Variations on a theme
So with the ability to talk to your phone and say “Phone Home” you’ve got some basic level of communication with your phone.
What’s more amazing is when it works! What’s irritating is when it doesn’t work.
The concept has been expanded to a range of devices for the home, mainly focused on entertainment.
Insert now the “virtual assistant”. A software agent that can perform tasks or services.
The Amazon Echo, Google Home, Playstation and XBox and Apple HomePod all offer a device than can listen to you and respond to commands.
Within those devices is a connection to the internet. It listens to what you say and transmits that audio to big servers. The big servers analyse the audio and do something rather clever.
The system starts by requiring a keyword to activate. “Alexa” in the case of Amazon’s Echo. This allows the device know what it hears next is for it to execute.
What is NLP?
NLP is an area of science which draws from computer science, data science and artificial intelligence.
It takes audio and works out what is being said. It turns the audio into text.
Then combining that text with historic instructions and “likely” statements it is able to work work out what traditionally you would have typed in.
The next big hurdle in the building block is being able to handle different languages. Then within each language you have dialects, accents and idioms.
Within Irish, there are three major dialects, Ulster, Munster and Connacht. Same language, three variations.
Yet there are hundreds of different accents within Ireland as well, all speaking English.
An idiom is defined as “a group of words established by usage as having a meaning not deducible from those of the individual words”. “Over the moon” means excited, not literally over the moon for example.
Multiple disciplines
NLP is building approaches to handle the variations.
Human brains handle the challenge very well and even humans have difficulty but using the brain’s approach to problem solving gives rise to how “artificial intelligence” is involved.
Having millions of examples to learn from is great. Data science is interested in being able to sift through all the data to find your idiom and deal with it correctly. Imagine how many examples Alexa will transmit back to the servers in its lifetime.
Putting all the pieces together and executing instructions is where the computer science comes in.
The long and the short of it is that this technology will work. The science and technical capability is there. Industry and science are just building the rules to help computers understand the audio input better.
Universities teach the science of Natural Language Processing.
Using the technology
Imagination from Science Fiction
For fans of the comic book superhero Iron Man, you’ll know that Tony Stark has a virtual assistant called JARVIS.
Jarvis is an acronym for for Just A Rather Very Intelligent System.
Now for the marvel movie nerds, Tony has a new virtual assistant called FRIDAY. With a very Irish accent from actress Kerry Condon which I find very cool.
Science fiction is inspiring science fact. So here’s what chatting to a natural language processing virtual assistant might look like.
Forgetting the visuals, what NLP is focused on is the audio. Augmented reality is a whole other discussion.
How interacting with a virtual assistant might be like. Note the call to Jarvis, like the call to Alexa, before each instruction.
You only need a the first few seconds to the get idea.
Applications beyond entertainment
What if you could chat to computer instead of a person? What could you do? Could NLP interpret?
- Call mum
- Pay my bills
- Order the groceries
- Add an item to the diary
- Record this program
- Turn the heating on
- Call emergency services
All of these are doable today!
Now instead of the person having a virtual assistant you could call one? i.e. the business you call has invested in a virtual assistant.
Speaking in tongues
Douglas Adams wrote a book called the Hitchhikers Guide to the Galaxy. In this science fiction / fantasy novel, creatures spoke many languages and many dialects. To help, a small fish, called a babel fish, swam in your ear and did the translation for you in your head.
Skype introduced an amazing feature.
- Firstly it listens to what you said and turns it into text.
- Next the text is translated by a translation engine, like google translate.
- This translated text is sent to a voice engine that verbalizes the text.
Suddenly with very little lag you have a real time translator. NLP needs to be working well to facilitate this.
Revolutionising customer care
What if your interaction with a business was?
- Person : Hi customer care, I’m having trouble with my bill?
- Virtual Assistant : No problem, can you just confirm your account?
Seeing how this changes the customer care industry becomes instantly understandable.
For all minor queries and standard requests, you could chat to an NLP enabled multilingual virtual assistant.
Contact centre teams will reduce to very advanced support teams only.
The escalation process over time becomes better and better. More and more queries are delivered by the virtual assistant.
The question becomes how you embrace this technology.
What can you do today
The first thing in doing to prepare is to centralise your business
Furthermore ensure your technology and business model is modular to be able to adapt and work with emerging solutions.
Also processes need to be upgrade to the point everything is as automated as possible.
Consequently ongoing learning and partnering with companies who are adapting to change is essential.
If haven’t got a data warehouse yet, now is the time to start working on one.
You don’t have to understand precisely how the technology works if you can get help from someone who does. The business basics of outsourcing.
[/read]
