Introduction

The NLU (Natural Language Understanding) Service is one of the main components in conversational systems (Fig. 1). Its function is to apply natural language understanding techniques to analyze user input sentences, allowing the dialog system to interpret them, extract information, and make decisions regarding the user interaction flow.

../_images/dialog-system.png

Fig. 1 Conversational artificial intelligence system.

The system is structured around the concept of an Agent which must be previously trained using data provided by the application developer. For example, for each intention, the developer needs to provide a list of example phrases from which the NLU Service will be able to identify that intention. Once trained, the Agent can perform text analysis. Based on a user’s input sentence, the NLU Service returns the intention and the identified entities (Fig. 2).

../_images/agent.png

Fig. 2 Sentence analysis

The integration with NLU Services should be done through the API REST which enables:

  • Agent management

  • Agent training

  • Sentence analysis in conversation

For an overview of how to use NLU Service, start with the User guide section.

For REST API details, see the API REST section.