StatusNeo

Google DialogFlow Part – 1

Introduction

• Dialogflow is used to create a tool that helps in interacting with your website, mobile app, or your product.
• It is a natural language understanding platform.
• It helps in creating a better user interaction tool so that the best interactive experiences can be provided.
• By providing input and output messages automatic responses can be generated using AI capabilities.
• Machine Learning algorithm helps in providing better response on basis of previous interactions.
• Dialogue in DialogFlow means the questions or statements that the user asks while interacting.

Dialog Matching

• We need to provide some sample questions and answers so that we can train our model for the same.
• Say for example we need to have a chatbot for our ShoeShop, so the sample questions will be like-
  • Is this(any size) available?
  • Is this color available for this size?
  • What is the delivery date?
  • How to proceed with the return?
  • When will this product come?
• Now, whenever these questions are asked by the user we have a text or audio base reply for this.
• And from the previous questions, we can train our model for similar questions like this.

Identifying User’s Expectations (Entities Extraction)

• Like in a database management system we have entities, similarly, we have the same in DialogFlow.
• Entities can be of different types like system entities, developer entities.
• In System Entities, we can have dates, times, emails.
• In Developer Entities, we can have new shoes, replacements, exchanges.
• We can also have synonyms of them so that a proper mapping can be done.
• Each word is made to fall into an entity so that classification can be done properly and better results can be achieved.

Control the dialog

• There are two types of dialog Linear dialog and non-linear dialog.
• The linear dialog includes one-liners questions and answers.
• That is questions having direct answers to them.
• Examples of linear dialogs are-
  • Can this be delivered to this place?
  • Is this size available?
  • Is this color available?
• While non-linear dialogs are the complex ones in which multiple answers are possible.
• Examples of non-linear dialogs are-
  • Can it be delivered tomorrow?
  • Two answers are possible for this – yes it is, no what about the day after tomorrow?
  • Now in the second case, there will be a series of questions.
  • This falls in the example of non-linear dialogs.

Conclusion

This is the first part of DialogFlow, in the next part we will cover the use cases, key features, and implementations.

References

• https://cloud.google.com/dialogflow
• https://www.youtube.com/watch?v=-tOamKtmxdY&list=PListM28kf0CNiMdmz-Ta5UcLR-ePCOW__&index=3

Still Curious? Visit my website to know more!

For more interesting Blogs Visit- Utkarsh Shukla Author