This section covers, essentially, the final two requirements of this API. We're going to create a JSON file that maps an intent to a possible reply (or replies). In fact, we're going to link each intent to multiple replies and then use one at random; that way, the bot will give the impression of responding uniquely to each message.
Let's start with the JSON document. Here's what it might look like:
{
"Hello": [
"Hi there!",
"Hello, My name is Boris",
"Nice to meet you",
"Hello"
],
"Family": [
"Tell me more about your family",
"Do you have any siblings?",
"Let's talk about your family some more"
],
"Feelings": [
"It's important to talk about your feelings",
"It's healthy to talk about how you feel",
"Please tell me: ...
