A machine has to grasp the natural language for it to be able to classify a user message according to what the user intends. The understanding of natural language rests on the previously fed data and the configuration of the NLU model. In case of a need to expand the model to deal with new use cases, model training must occur. Our Training Feature is available to support this process.
The Training Feature employs artificial intelligence to train an NLU model based on data, such as intents, phrases, and entities in the NLU module. By adding or editing this data, the feature trains the model to learn and store this new information to accurately detect future intents, phrases, and entities. In other words, by training our model in an iterative and appropriate* manner, phrases, intents, and entities are verified, compiled, and stored to the model for future use and further bot development, enabling the bot to improve and understand more over time.
Let’s take a closer look at the training process using Automate’s Train button.
To train the NLU, click the Train button in the upper-right corner of the Automate interface.
These are three statuses of training:
Updated about 1 year ago