Chatbot Common Terminology
Chatbot or intelligent bots or just bots, you may call them whatever you wish to. Gradually, they are becoming an integral part of our daily life. Because of its capabilities to get integrated with almost any messaging service say it WhatsApp, Facebook messenger, Snapchat, etc. they are becoming more and more known to us.
As many of us want to learn about chatbots and the way they function, I thought it would be good to share the common terminology used in chatbot development. So, here we go.
Intent: It's a unit of work you intend to perform using the bot. It could be ordering pizza, checking your bank account balance, reserving a seat in your favourite restaurant, etc.
Utterance: Example phrases, to understand the Intent, is called utterance. Bot will then use AI to able to understand the phrase. For each Intent we generally train the bots with 10 to 15 Utterances. Bot should still detect phrases which are not part of Utterances and respond to them as it is unknown to them.
Entity: As the definition goes, it is the variable element of Intent that we intend to Parameterize to allow input of different Values.
In other words, its the input value(s) in the a sentence while communicating with bot, that we pass to the bot, based on which we seek outcome(s). You can use this value then, to may be invoking a REST call to fetch the result.
A bot will first resolve the Intent and then resolve the Entity associated to that Intent.
Intelligence: When interacting with a bot, we type sentence without being aware of the exact phrase to be passed. A rule based approach simply won't work in case of bots.
For example, we might have defined an Intent as "Order a pizza". But without being aware of this, an end user may enter a text as "I'm hungry. Order something to eat, may be pizza".
At this time Artificial Intelligence(AI) comes into the picture to understand the Intent of user. This is done by the Natural-language processing (NLP) of AI engine.
Flow: Interaction with a user and a bot compose of a series of word exchange, in a specific manner, is known as a Flow.
In other words, steps followed by user and a bot during their conversation constitutes a flow. Let's take an example of ordering a pizza.
A bot will have a series of conversation with an user before ordering the pizza.. Like which type of pizza, what will be the size of pizza, any toppings over that, any side dish, etc.
This whole conversation take place as a flow, in a sequential manner.
As many of us want to learn about chatbots and the way they function, I thought it would be good to share the common terminology used in chatbot development. So, here we go.
Intent: It's a unit of work you intend to perform using the bot. It could be ordering pizza, checking your bank account balance, reserving a seat in your favourite restaurant, etc.
Utterance: Example phrases, to understand the Intent, is called utterance. Bot will then use AI to able to understand the phrase. For each Intent we generally train the bots with 10 to 15 Utterances. Bot should still detect phrases which are not part of Utterances and respond to them as it is unknown to them.
Entity: As the definition goes, it is the variable element of Intent that we intend to Parameterize to allow input of different Values.
In other words, its the input value(s) in the a sentence while communicating with bot, that we pass to the bot, based on which we seek outcome(s). You can use this value then, to may be invoking a REST call to fetch the result.
A bot will first resolve the Intent and then resolve the Entity associated to that Intent.
Intelligence: When interacting with a bot, we type sentence without being aware of the exact phrase to be passed. A rule based approach simply won't work in case of bots.
For example, we might have defined an Intent as "Order a pizza". But without being aware of this, an end user may enter a text as "I'm hungry. Order something to eat, may be pizza".
At this time Artificial Intelligence(AI) comes into the picture to understand the Intent of user. This is done by the Natural-language processing (NLP) of AI engine.
Flow: Interaction with a user and a bot compose of a series of word exchange, in a specific manner, is known as a Flow.
In other words, steps followed by user and a bot during their conversation constitutes a flow. Let's take an example of ordering a pizza.
A bot will have a series of conversation with an user before ordering the pizza.. Like which type of pizza, what will be the size of pizza, any toppings over that, any side dish, etc.
This whole conversation take place as a flow, in a sequential manner.
Comments
Post a Comment