What are Entities in Dialogflow? Everything you need to know about the 3 yypes of Entities

#1. What is an entity in Dialogflow?

An entity in Dialogflow is a mechanism that helps recognize and extract specific data from user input (text or voice).

It allows the agent (chatbot representative) to retrieve detailed information such as dates, times, locations, phone numbers, or other custom data defined by users.

For example, if a user says, "Show me the address of restaurant X in city H," the entity in this case will be:

• Location/Restaurant: The name of the restaurant, which is X. This entity can be stored as @restaurant_name.
• City: The name of the city, which is H. This entity can be stored as @city_name.

When extracting entities, the results would be:

• @restaurant_name: X
• @city_name: H

In Dialogflow, entity names begin with the @ symbol to differentiate them from other components such as intents (user intentions) or actions (agent actions).

Depending on the context, you can define your own entities, such as car types, music genres, quantities, or units. Any important data you want to extract from user requests should have a corresponding entity.

#2. Types of entities in Dialogflow

In Dialogflow, there are three main types of entities:

#System entities

System entities are pre-built entities in Dialogflow that allow agents to extract information about various concepts without additional configuration.

For example, Dialogflow provides default system entities to extract dates, times, and locations from natural language input.

Example:

If a user says: "I want to book a table for next Friday at 7 PM." The extracted entities would be:

• @sys.date: "next Friday"
• @sys.time: "7 PM"

Both of these are system entities, meaning they are pre-built and require no additional setup.

You should use developer entities when you need to recognize specific words or phrases that are not available in system entities, or if you want more control over entity recognition and processing.

#Developer entities

Developer entities are custom entities defined by developers to meet specific chatbot requirements.

Example:

If you are building a movie ticket booking chatbot, you can create the following developer entities:

• @movie_name: A list of movie names. Values: ["Avengers", "Inception", "Titanic"]
• @cinema_name: A list of cinema names. Values: ["Galaxy", "CGV", "Lotte"]
• @seat_type: Types of seats in the cinema. Values: ["Regular", "VIP", "Recliner"]

When a user says: "I want to book 2 tickets for Avengers at CGV in VIP seats," Dialogflow can extract the following entities:

• @movie_name: "Avengers"
• @cinema_name: "CGV"
• @seat_type: "VIP"

#Session entities

Each conversation between a Dialogflow agent and an end user is considered a session. Therefore, you can create special entities that exist only within a single session, known as session entities.

Simply put, session entities are temporary versions of entities, defined specifically for a particular conversation session.

Example:

Suppose you have a chatbot for movie ticket booking. You have a developer entity called @movie_name with three popular movies:

• @movie_name: ["Avengers", "Inception", "Titanic"]

However, if a user asks about a special movie event happening today, you can create a session entity to include these special movies.

At this point, you will have a session entity also named @movie_name, but with different values: ["Special Movie 1", "Special Movie 2"].

If the user says: "I want to book tickets for Special Movie 1," the chatbot can recognize @movie_name as "Special Movie 1."

This process does not affect the original developer entity, which still contains only three values.

Session entities can extend or override custom entities and exist only during the session in which they are created. Once the session ends, session entities are deleted and do not impact fixed entities. All session data, including session entities, is stored in Dialogflow for up to 20 minutes.

#3. Conclusion

As you can see, entities play a crucial role in Dialogflow, helping chatbots understand and extract important information from user input. Besides entities, you should also explore intents, agents, and training phrases to master Dialogflow effectively!

[EmbedSoftware ids="54"]