NLEinnd—QuickNLP interpretation end
Logged by the Natural Language Engine just after the interpretation call to the core engine (QuickNLP) has completed. Includes the output of the core engine.
There is one such event per Interpret command.
In addition to the Tokens used for every event, this event has the following tokens:
Token | Meaning |
---|---|
INTENT(0) | Intention of the top n-best result. |
KEYS(x) | XML formatted result with all key-value pairs, iterated per acoustic (n-best) result. See below for the list of keys. |
NLCONF(0) |
The recognition engine's confidence in the intent assigned. Values can range from 0 to 99. |
RAWT(x) | Raw text for n-best item. Pre-tokenized minimally-formatted form (MFF). |
Note: The KEYS(x) and RAWT(x) token sequence will appear once for each n-best result returned. These tokens do not appear when there are no associated results.
NLEinnd keys
Key | Meaning |
---|---|
INTENT | Intention assigned to the input. For example, BILL_PAY. |
<ENTITY_NAME> | Value of the given entity name, where the name is the entity (for example <PAYEE>) and the value is the entity value (for example, VISA). See the examples below. |
NLCONF |
The recognition engine's confidence in the intent assigned. Values can range from 0 to 99. |
WHENCE | Origin of the engine result. |
The KEYS token is in XML format. The format is flat XML without nesting. For example:
<INTENT>BILL_PAY<INTENT/><PAYEE>VISA<PAYEE/><AMOUNT>255.00<AMOUNT/><NLCONF>99<NLCONF/><WHENCE>SSM<WHENCE/>