<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=979905748791482&amp;ev=PageView&amp;noscript=1">

Natural language processing

Dig a little deeper with a human-friendly look at the technology powering world-class chatbots.
Talk to chatbot experts

What is natural language processing?

Natural language processing (NLP) is the technology that enables a computer to translate human-speak into a language it recognises. NLP is a subset of artificial intelligence. Chatbots use natural language processing to understand human input and respond accordingly.

Talk to our experts

NLP in the real world

The technology traces its roots back to the 1950s, with Alan Turing's famous 'Computing Machinery and Intelligence' article. Today, NLP is everywhere. You most likely have been using it without realising.

NLP can be found in your Alexa or Siri's speech recognition, the auto-generated search recommendations you spot on Google, and word-generation and spell checking in that news article you just read. Of course, chatbots make great use of this technology, too.

What about natural language understanding?

If natural language processing is the building of computer logic and programs based on human-language, then natural language understanding (NLU) is all about enabling a computer to comprehend what the human input means. It's a subtopic of natural language processing, much narrower in purpose. Rarely, if ever, does natural language processing not need any prior understanding.

Imagine natural language understanding as the work-horse; it's doing the stuff in the background for the computer to communicate with you effectively. We use this technology for text categorisation, understanding millions of news articles, rerouting emails or even comprehending what options a new vehicle is sold with.

Common uses of NLP in chatbots

Who creates (and maintains) a chatbot's NLP?

Different parties can 'power' the NLP of your chatbot, and you may not always know where your data is going.
Some chatbot development companies rely on (and resell) NLP services from big tech companies like Google, Amazon, or Facebook. While this can speed up development, it also means all the data your chatbot gathers goes straight to the aforementioned tech giants. Good luck getting that through your legal department.
Some chatbot development companies (like ubisend) have their own NLP engine, especially programmed and trained on what you need your chatbot to achieve. What this approach lacks in sheer amount of data, it more than makes up for in preciseness. By going niche and only training a chatbot to know what you need it to understand, you create a highly-efficient machine.
This approach also means you know precisely where your valuable data is going.
creating chatbot nlp engine
chatbot nlp training management suite

The ongoing process of training a chatbot

Imagine your new chatbot as a baby. It's fresh out of early development and only had a few hundred test conversations. Your chatbot doesn't know what it is, it doesn't know who you are, and it certainly doesn't understand your complex HR handbook or frequent customer service questions. Your chatbot needs to learn; it needs to be trained.
We've built many chatbots, and are able to speed up the growth of your chatbot. Through our in-house technologies and vast data sets, we usually ship your chatbot to you as a teenager (don't worry, not as grumpy as a human one). It's trained to know who you are and the content you need it to understand. But it still has a lot to learn.
This is where chatbot training comes into play. It's up to you, with our help, to keep educating and improving your chatbot. Through correcting and reinforcing its NLP, your chatbot becomes more and more apt. Don't worry; you don't need a team of data scientists; it's all done through a human-friendly administration software suite we ship with your custom chatbot.

"It's been great to work with ubisend, it is clear that they are a driven and determined team of professionals who provided us with the top quality solution we needed where initially the end product was far from clear."

Chris Amos, project manager at Archant Limited

We've built solutions for many companies, here are a few.

Latest Insights