Utiliza modelos prediseñados

How to use pre-designed models in LUIS to enhance applications?

Incorporating pre-designed models in your LUIS (Language Understanding Intelligent Service) applications is an efficient way to add common functionalities without having to develop them from scratch. These models offer preconfigured entities and intents that can significantly speed up the development process, especially when working with applications that need to process natural language in certain areas, such as weather or e-mail.

What aspects to consider when choosing a pre-designed model?

• Language and regional variations: When creating an application in LUIS, it is crucial to define the language correctly. Some regional settings do not support prebuilt templates. For example, for Mexican Spanish, there are no prebuilt domains available. A good practice is to opt for neutral Spanish if you want to use these features.

• Model configuration: You can start by creating a new application (e.g. "travel app") in neutral Spanish. Select the appropriate prediction services and explore the pre-built domains you can add to your app.

How to add pre-built domains?

To add pre-built domains, follow these steps:

1. Navigate to the "Previewint" menu and select "Pre Build Domains".
2. Review the available domains, which may include features such as email management or weather queries.
3. Select the domain that best fits your application's needs.

What do the pre-built intents and entities offer?

The pre-built domains provide you with examples and variant phrases to feed your intentions. Note that:

• Intents: There are multiple examples for each intent, which reduces manual work to a minimum.
• Entities: Include pre-built data, such as historical weather and temperature ranges, which are automatically integrated into your application.

What to do if your regional variant does not support pre-built domains?

In case your regional configuration does not support prebuilt models, you can still use prebuilt entities to facilitate development. These include:

• Age
• Dimensions
• Email addresses
• Numbers and percentages
• Temperatures and other common data

For example, by adding a pre-built entity such as "temperature", you will be able to train the model and associate new examples to connect this entity with your intentions.

Best practices to speed up development with pre-built models

• Time optimization: Use pre-built domains and entities to drastically reduce your application setup time.
• Adaptability: Although Pre Build Domains may not support all regional configurations, you can opt for pre-built entities for common fields.
• Continuous training: Once you add entities and intents, train your model to improve its accuracy and efficiency in interpreting inputs.

Using pre-designed models and entities in LUIS not only streamlines your work, but also facilitates the integration of highly sophisticated features for natural language processing. By understanding the options available and how to implement them, you can create smarter and more robust applications. Keep exploring and learning, your journey into the world of natural language and intelligent applications is just beginning!