I see that the slot mapping for KB is done internally in RASA. I used InMemoryKB for the task. It is a simple ticket enquiry system. My question is when we have multiple tickets in the same utterance how are the slots filled. As far as i saw, only one value gets mapped coz i used the type text. How doe the list slot work with KBs. Can someone share any reference?
2020-10-15 10:48:00 DEBUG rasa.core.processor - Received user message ‘what is the status of 123 and 225’ with intent ‘{‘id’: -1636562065130203629, ‘name’: ‘ticket_info’, ‘confidence’: 0.9996767640113831}’ and entities ‘[{‘entity’: ‘attribute’, ‘start’: 12, ‘end’: 18, ‘confidence_entity’: 0.9993613362312317, ‘value’: ‘status’, ‘extractor’: ‘DIETClassifier’}, {‘entity’: ‘ticket’, ‘start’: 22, ‘end’: 25, ‘confidence_entity’: 0.9995343685150146, ‘value’: ‘123’, ‘extractor’: ‘DIETClassifier’}, {‘entity’: ‘ticket’, ‘start’: 30, ‘end’: 33, ‘confidence_entity’: 0.9996114373207092, ‘value’: ‘225’, ‘extractor’: ‘DIETClassifier’}, {‘start’: 22, ‘end’: 25, ‘text’: ‘123’, ‘value’: 123, ‘confidence’: 1.0, ‘additional_info’: {‘value’: 123, ‘type’: ‘value’}, ‘entity’: ‘number’, ‘extractor’: ‘DucklingEntityExtractor’}, {‘start’: 30, ‘end’: 33, ‘text’: ‘225’, ‘value’: 225, ‘confidence’: 1.0, ‘additional_info’: {‘value’: 225, ‘type’: ‘value’}, ‘entity’: ‘number’, ‘extractor’: ‘DucklingEntityExtractor’}]’
2020-10-15 10:48:00 DEBUG rasa.core.processor - Current slot values:
attribute: status
email: None
knowledge_base_last_object: None
knowledge_base_last_object_type: None
knowledge_base_listed_objects: None
mention: None
num_emails: None
object_type: None
plan: None
priority: None
requested_slot: None
status: None
ticket: 225
topic: None