Please enable JavaScript.
Coggle requires JavaScript to display documents.
Chatbot - Coggle Diagram
Chatbot
Building a chatbot
Create agent
Name the intent
Add training phases
Add training phase
Add response for intent
Create follow-up intents
Define entities
Use unique token
Do not share token with unauthorized personnel
Test and train the Chatbot
Chatbot is used for automated online chat conversation instead of using a human agent
Chatbot uses Natural Language Processing (NLP) to decipher human language
Chatbots provide accurate answers quickly
Chatbots provide replies to customers anytime of the day or night
Chatbots relieve human agents of workload
Chatbots log conversation history to analyse needs of customers
Chatbots are built on different platforms like Google Dialogflow and IBM Watson Assistant
There are three concepts in Google Dialogflow - Agent, Intent and Entity
AGENT
Doesn't need precise grammar and syntax
Handles end-to-end conversations with users
Training is needed
INTENT
Identifies the customer's intent for a conversation turn
Chatbot needs to learn about end-user phrases to categorise intents
ENTITY
It is a specific data or perimeter to help better make sense of what the user is trying to find out
e.g dates, times, numbers and email addresses