When user answers as “I would like to speak in Spanish” to the question “Which language would like to proceed with”.
For the AI Agent to proceed further only needs to know the keyword “Spanish” and not the whole sentence.
To get the only required keyword from a sentence, Entity are used.
Types of Entity
There are 2 types of Entity.
Custom entity: Which needs to be created by the user as per the use case.
System Entity: Some of the predefined entities like Language, date, email, duration are available.
Adding Entity and adding synonyms
Creating an Entity includes, Naming the Entity, adding synonyms, adding reference value. Please refer to the video below to check out more on how to create an Entity.
Marking Entity to the intents
Once the entities are added, now to use them we need to mark the entity to the respective Intents.
For e.g.: You have created an entity for language with reference value of “Spanish” and “English”. The variations which I have is “I would like to speak in Spanish”. So, to decide the workflow we only need the keyword “Spanish” and not the full sentence. So, we will go and mark the keyword of language to the multiple variations. Please refer to the video below for more details.
As you can see in the above video, we had an entity named Language. We marked the keyword “English” and “Spanish” to the variations. At the bottom of all the variations you can see a table named “Parameters” which has entity as “language” and the value as “language”.
By marking this, When I input an answer by saying “I want to speak in Spanish”, the Entity will resolve the sentence and will set the value as “Spanish” to the value “language”.
Using Entity in workflow
Please refer to the below video to know the importance of Entity and how it will be helpful to get accurate answers.
