Language Understanding (LUIS)
Language Understanding (LUIS) is a cloud-based service that applies custom natural language understanding models. LUIS enables you to build applications that can understand what users want by analyzing their utterances. It extracts intent and entities from natural language text.
LUIS allows you to create custom models to understand the specific language of your users, no matter the domain. This makes it ideal for building conversational interfaces, chatbots, and other AI-powered applications.
Get Started with LUISOn This Page
What is LUIS?
Language Understanding (LUIS) is a part of Azure Cognitive Services that allows developers to create custom natural language understanding (NLU) models. It helps applications understand user input by identifying intents and extracting key information (entities).
- Intents: The user's goal or action they want to perform (e.g., "BookFlight", "CheckWeather").
- Entities: Specific pieces of information within the user's utterance that are relevant to the intent (e.g., "New York" as a location, "tomorrow" as a date).
By training LUIS with example utterances, you can build a powerful NLU engine tailored to your specific application's needs.
Key Concepts
Utterances
Utterances are the sample phrases users might say or type to your application. The more diverse and representative your utterances are, the better your LUIS model will perform.
Example: "Book a flight to London for me and my family tomorrow."
Intents
An intent represents an action or a goal a user wants to achieve. You define intents that your application should recognize.
Example Intents: BookFlight, CancelOrder, GetDirections.
Entities
Entities are the specific pieces of information within an utterance that LUIS extracts. They are used to parameterize intents or provide data for your application.
Example Entities for BookFlight intent:
Destination(e.g., "London")Date(e.g., "tomorrow")NumberOfPassengers(e.g., "me and my family" could be inferred or extracted as a number)
Domains
LUIS offers pre-built domains that provide starter models for common scenarios, such as "Home Automation," "Conversational Business," or "Order Status." These can significantly speed up development.
How LUIS Works
- Create an app: Define intents and entities relevant to your application.
- Add utterances: Provide example phrases for each intent.
- Label entities: Tag specific words or phrases in utterances as entities.
- Train the model: LUIS uses your data to train a machine learning model.
- Test and improve: Evaluate the model's performance and iteratively add more data or refine existing definitions.
- Publish: Make your trained model available via an endpoint to be consumed by your application.
Tutorials and Quickstarts
Dive into building your first LUIS application with our comprehensive guides:
- Quickstart: Create your first LUIS app
- Tutorial: Build a chatbot that understands intents
- How to manage entities in LUIS
- Best practices for LUIS app development
SDK and API Reference
Integrate LUIS into your applications using our robust SDKs and REST APIs.
Pricing and Limits
Explore the pricing tiers for LUIS and understand the usage limits associated with each plan.
Support and Community
Get help and connect with other developers: