Please enable JavaScript.
Coggle requires JavaScript to display documents.
Chatbots - Coggle Diagram
Chatbots
Why?
Trained chatbots can provide accurate answers quickly, whereas a human agent may need time to retrieve information and its subject to human errors
Chatbots, whicg are automated, can provide replies to customers anytime of the day (or night), ensuring a satisfied customer at anytime of the day (or night)
Chatbots - being bots - can relieve human agents of workload; human agents can focus on other tasks that cannot be automated
Chatnots are able to log conversation history, which can then used to analyse needs of customers
Google Dialogflow
AGENT
-
Needs to be trained, although not explicity (i.e. don't need to train with exact words and phrases)
-
INTENT
-
When the user says something, it is referred to as the "end-user expression"
Unlike human beings who have learnt and gain experience, the chatbot needs to learn about end-user phrases to categorise intents
E.g. "Whattime do you open on Saturday?" and " What are the opening hours on weekends?" have the same intent
ENTITY
Entity is specific data or parameter. They are structures data that can help us make better sense of what the user is trying to find out
Examples are dates, times, number, email addrasses
Possible to create your own custom entities, e.g define a "vegetable" entiy and populate it with a list of vegetables
How dialogflow work
- Agent receives an end-user expression
- Agent attempts to match the end-user expression to what it believes in the intent of the user asking the questions
- Once agent matches the intent, the response is sent back to the user
Improvements to Chatbot
-
Requires trainning and iterations to improve the chatbot such that it can provide accurate information everytime
-