Intent matching based on the context

[Brindha]

Hi @xavidop ,
  In dialogflow cx, how can we match the intents not only with the training phrases we added but also analyzing the context of the user prompt?
For example, Here I'm having two scenarios like 1) login issue and 2) project access in some platform. 
I added basic training phrases for both.
For login issue--> "I'm trouble login to jira", "login issue"
For project access --> "I need access for the project", "Please add me to the project"

If the user says, "I cannot access jira. Could you please help me?"
This gets into project access flow. But the context related to login issue flow.
How can we match intent based on the context?

Thanks in advance

[paulynann]

Hi @Brindha,

Welcome to Google Cloud Community!

To effectively match intent based on context, Dialogflow CX uses certain methods. This includes:

• Entities: This represents specific information within an end user’s query. Dialogflow CX can extract entities from the end user’s query using a variety of techniques, including pattern matching, machine learning, and natural language processing. Entities can be used to identify things like names, dates, locations, and more. Once identified, entities can be used to personalize the conversation and provide more relevant responses. For example, if a user mentions a specific product like Jira, Dialogflow CX can use that entity to provide information about that product or make recommendations.
• Parameters: This is used to capture and reference values that have been supplied by the end-user during a conversation. Unlike raw end-user input, parameters are structured data with a name and an entity type. This structure allows them to be easily used for performing logical operations or generating responses. For example, if a user mentions Jira, Dialogflow CX can store that information as a parameter and use it to generate a personalized response.
• Fulfillments: This allows you to use the information extracted by Dialogflow's natural-language processing to generate dynamic responses or trigger actions on your backend. For example, you can use a fulfillment to send a message to a user, like providing instructions on how to reset the user's password. Fulfillments are a powerful tool that can be used to extend the capabilities of Dialogflow CX and create more engaging and informative conversations.

By combining these methods, Dialogflow CX is able to understand the user's intent and provide the most relevant response. This results in a more natural and engaging conversation experience for the user.

In addition, I recommend following the best practices for Dialogflow CX as well.

I hope the above information is helpful.

[Mizar]

Hi @Brindha for this case I recommend you to create entities for Login Issue (that can cover, I cannot access, etc) and also Jira as another entity
Then if the user says I cannot access Jira, you can trigger when someone says I cannot access Jira.

Then you can create an entity from code called context and you can save a value you want so then you can go on to the next intent and use this context parameter to nurture the best response.