Automotive team is working on developing a functionality in our small talk skill to identify names. The skill uses regexp engine, and we have defined a custom entity called names with 400+ names.
The issue we're having is that the utterance "Hi my name is Mary Anne" only resolves 1 name which is "Mary". The request object in the skill does not have "Anne", which it should.
When we flipped the names to "Hi my name is Anne Mary", only "Anne" is in the request object. Is this the expected behavior of the regexp NLU engine? If so, could the utterance be processed completely for each entity type and all occurrences be return? @offerakrabi
Automotive team is working on developing a functionality in our small talk skill to identify names. The skill uses regexp engine, and we have defined a custom entity called names with 400+ names.
The issue we're having is that the utterance "Hi my name is Mary Anne" only resolves 1 name which is "Mary". The request object in the skill does not have "Anne", which it should.
When we flipped the names to "Hi my name is Anne Mary", only "Anne" is in the request object. Is this the expected behavior of the regexp NLU engine? If so, could the utterance be processed completely for each entity type and all occurrences be return? @offerakrabi