hi,I wanted to know what Machine Learning algorithm used to identify intent and entities from user input sentence.
1 Like
It depends on what pipeline you chose to process your data (see Choosing a Pipeline).
But in general, we use a CRF model to extract the entities (see documentation) and for the intent classification we rely on StarSpace (see documentation). You can read everything about the individual components here: Components.
2 Likes
Thank you, will look into it.
Hi, is this approach still relevant in the newest version (3.0.x)? (for default recipe NLU pipeline)