System and method for a natural language understanding system based on iterative intent detection and slot filling neural layers

A natural language system, comprising:

• a processor programmed to:
• receive data indicative of a natural language statement;
• provide one or more tokens associated with the natural language statement to a recurrent neural network;
• output one or more intent representation vectors in response to the one or more tokens and output one or more slot representation vectors in response to both the one or more tokens and the one or more intent representation vectors during a first iteration;
• and
• utilizing at least a second iteration in addition to the first iteration, output a sentence intent and a slot label based on the one or more tokens, the one or more intent representation vectors, or the one or more slot representation vectors from the first iteration.

A natural language system, comprising:

• a processor programmed to:
• receive data indicative of a natural language statement;
• provide one or more word embedding vectors associated with the natural language statement to a recurrent neural network;
• output one or more intent representation vectors in response to the one or more word embedding vectors and output one or more slot representation vectors in response to both the one or more word embedding vectors and the one or more intent representation vectors during a first iteration;
• and
• utilizing at least one or more high-level iterations in addition to the first iteration, output a sentence intent and a slot label based on the one or more word embedding vectors, the one or more intent representation vectors, or the one or more slot representation vectors from a previous iteration.