Let’s imagine a user enters a chat room and says "Hey, guys". This is a common introductory sentence in almost every chatroom. Suddenly, a chatbot responds, "We are committed to diversity, so let’s use a more gender-neutral greeting". This can easily done by using a kind of regex.
But, what if that sentence is a little different? "Guys" has a lot of uses. Maybe this sentence is "Lots of guys are leaving their homes to go to a soccer game". In this case, it still needs the diversity reminder, as women can do the same as men. However, the bot needs to react differently, as this use case is not a greeting. This is where NLU, or Natural Language Understanding, comes on to the scene.
In this session, we’ll go over the basics of natural language understanding as well as the data security implications in using them. After all, we all know and use Alexa, but is our data safe? And is it even possible to create an open source chatbot without running into privacy concerns?