Discover Exciting Language Understanding Advancements

Table of Contents

1. Introduction
2. What is Language Understanding?
3. New Features in Language Understanding
   • Updated Portal Experience
   • Deeper Hierarchy for Entities
   • Seamlessly Upgrading Old Composite Entities
   • Normalized Word Forms
   • Docker Container Support
   • Batch Testing through the API
   • Enhanced Data Encryption
   • DevOps Sample Project
4. Using Language Understanding in Virtual Agents and Bots
5. Demo: Processing Leave Requests in a Bot
6. Understanding Intent and Entities in Language Understanding
   • Intent Recognition
   • Entity Recognition
7. Customizing and Labeling Utterances in the Portal
   • Improved Entity Labeling Experience
   • Labeling with the Custom Entity Palette
   • Inline Labeling
8. Conclusion
9. Additional Resources and Community

New Features in Language Understanding: Making AI More Accessible and Powerful

The field of language understanding has seen tremendous advancements in recent years, with the development of AI services that allow developers to integrate natural language understanding into their applications. One such service is Language Understanding (LUIS), which enables developers to easily extract intents and entities from text, allowing applications to understand user requests and process them accordingly.

In this article, we will explore the new features in Language Understanding that aim to make it even more powerful and accessible for developers. We will discuss the updated portal experience, the introduction of deeper hierarchy for entities, the seamless upgrade of old composite entities, the new app setting for normalized word forms, and much more.

Updated Portal Experience

One of the key improvements in Language Understanding is the updated portal experience. The new portal makes it easier than ever before to label utterances, update entity schema, and Create features. With a user-friendly interface and intuitive navigation, developers can quickly build and customize their language models.

Deeper Hierarchy for Entities

An exciting new feature in Language Understanding is the ability to add deeper hierarchy to machine-learned entities. This allows developers to recognize more complex intents and entities and reuse them across their applications. With the power of these new entities, developers can enhance the accuracy and efficiency of their language models.

Seamlessly Upgrading Old Composite Entities

To ensure a smooth transition for developers, Language Understanding offers seamless upgrades for old composite entities. If You have old composite entities in your application, the updated portal will notify you and provide an option to upgrade them. This process is completely risk-free, as it creates a new version of your app that can be thoroughly tested before deployment.

Normalized Word Forms

Language Understanding introduces a new app setting called normalized word forms. By default, LUIS treats root words differently from their plurals or words with suffixes. However, in certain cases, you may want the plurals and suffixes to be handled the same way in your application. With the normalized word forms setting, you can ensure consistency and simplify the processing of such cases.

Docker Container Support

Responding to customer demands, Language Understanding now offers Docker container support. This feature allows developers to run their language models anywhere and Scale them as needed. With Docker container support, there are no limitations on transactions per Second (TPS) when hosting the container, providing developers with maximum flexibility and scalability.

Batch Testing through the API

Language Understanding also introduces the ability to perform batch testing through the API. Previously, batch testing was only available in the portal, but now developers can conveniently test their language models using the API. This feature streamlines the testing process and allows for more efficient development and deployment.

Enhanced Data Encryption

In today's data-driven world, security is of utmost importance. Language Understanding now offers enhanced data encryption features to ensure the safety of your sensitive information. You can encrypt all the data you use in LUIS using your own key. This can be achieved by creating your own keys and storing them in a key vault, or by utilizing the Azure Key Vault APIs to generate keys. This enhanced security feature gives you full control over your data and provides peace of mind.

DevOps Sample Project

To facilitate the integration of Language Understanding into your development workflow, a DevOps sample project has been released. This template repository provides a working project structure and GitHub action pipelines, allowing you to customize it according to your own project requirements. With this DevOps sample project, you can streamline your development process and ensure seamless collaboration with your team.

These new features in Language Understanding are designed to make AI more accessible and powerful for developers. With an updated portal experience, deeper hierarchy for entities, seamless upgrades, normalized word forms, Docker container support, batch testing, enhanced data encryption, and the DevOps sample project, developers can create more accurate and efficient language models.

In the next section, we will explore how these new features can be used in virtual agents and bots to enhance customer support and streamline various processes.

Continue Reading

Demos Planned at Build 2021

At Build 2021, Language Understanding unveiled exciting new features and tools that are set to revolutionize the development of natural language processing models. Alicia Edelman Pelton, a Principal Program Manager in the Language Understanding Group and Cognitive Services, showcased the various advancements and provided a live demo to demonstrate the capabilities of these new features.

One of the highlights of these new features is the updated portal experience, which offers a more intuitive and efficient way to build and customize language models. The portal now allows developers to easily label utterances, update entity schema, and create features.

Another significant enhancement is the introduction of deeper hierarchy for entities. This feature enables developers to recognize more complex intents and entities, allowing for more accurate and robust language models. It also facilitates the reuse of entities across different applications, streamlining the development process.

Moreover, Language Understanding now offers seamless upgrades for old composite entities. This means that developers can upgrade to the latest version of the platform without the hassle of manually migrating their existing entities. The upgrade process is risk-free and ensures a smooth transition for developers.

In addition to these improvements, Language Understanding now supports normalized word forms. Developers can now handle root words, plurals, and words with suffixes in a consistent manner, making their applications more intuitive and user-friendly.

Furthermore, Language Understanding introduces Docker container support, enabling developers to run their language models anywhere and easily scale their applications. This provides greater flexibility and efficiency in deploying and managing language understanding models.

Another exciting feature announced at Build is the ability to perform batch testing through the API. Developers can now test their language models more efficiently and in large volumes, thanks to this new capability. This feature streamlines the testing process and enables developers to iterate quickly on their models.

In terms of security, Language Understanding now offers enhanced data encryption options. Developers can encrypt all the data they use in LUIS using their own keys, ensuring maximum security and compliance with data protection regulations.

Lastly, Language Understanding has partnered with the CSE team to release a DevOps sample project. This project provides a template repository that includes a working project structure and GitHub action pipelines. Developers can leverage this sample project to streamline their development workflow and facilitate collaboration within their teams.

With these new features and tools, developers can create more accurate and powerful language understanding models, allowing them to build intelligent applications that can better understand user intents and respond accordingly.

To learn more about these new features, you can refer to the official documentation on Language Understanding. Additional resources and community support are also available for those looking to further explore and leverage these advancements.

Continue Reading

FAQ

Q: Can Language Understanding be used in virtual agents and bots?

Yes, Language Understanding can be integrated into virtual agents and bots to enhance their natural language understanding capabilities. By extracting intents and entities from user queries, Language Understanding enables virtual agents and bots to understand user requests and respond accordingly.

Q: How does the new hierarchy for entities improve language models?

The new hierarchy for entities in Language Understanding allows developers to recognize more complex intents and entities. This enables the creation of language models that can understand and respond to a wider range of user queries with greater accuracy.

Q: Can old composite entities be seamlessly upgraded to the latest version?

Yes, Language Understanding offers seamless upgrades for old composite entities. Developers can upgrade their existing composite entities to the latest version without manual migration. The upgrade process is risk-free and ensures a smooth transition.

Q: What is the purpose of normalized word forms in Language Understanding?

Normalized word forms in Language Understanding allow developers to handle root words, plurals, and words with suffixes in a consistent manner. This feature makes applications more intuitive and user-friendly by treating related word forms in a unified way.

Q: How does Language Understanding support Docker containers?

Language Understanding now offers Docker container support, enabling developers to run their language models anywhere and easily scale their applications. This provides greater flexibility and efficiency in deploying and managing language understanding models.

Q: Can batch testing be performed through the Language Understanding API?

Yes, Language Understanding now supports batch testing through the API. Developers can efficiently test their language models in large volumes, streamlining the testing process and enabling quick iterations and improvements.

Q: How can developers ensure the security of their data in Language Understanding?

Language Understanding offers enhanced data encryption features, allowing developers to encrypt all the data they use in LUIS using their own keys. This ensures the security and privacy of sensitive information.

Q: What is the DevOps sample project in Language Understanding?

The DevOps sample project is a template repository provided by Language Understanding. It includes a working project structure and GitHub action pipelines, making it easier for developers to customize and integrate Language Understanding into their own projects.

Q: Where can I find more information and resources on Language Understanding?

For more information and resources on Language Understanding, you can refer to the official documentation. Additionally, the Language Understanding community is a valuable resource for support and further exploration of the platform.