knowledge_chunks

A list containing the top-20 matching knowledge chunks found in the knowledge base. Each knowledge chunk has the following fields:

matching_text - This is the text associated with the knowledge chunk which is used to match against the users query. In the default configuration, the matching_text is in the form of a question.

qualifying_text - This is a bit of text that provides contextual information needed to determine if this knowledge chunk is relevant or not. In the default configuration, the qualifying_text is just a summary of the document where the knowledge came from.

content_text - This is the main text of the knowledge chunk that contains the knowledge we retrieved. This is usually some snippet of text that was isolated and separated from the larger document it was apart of.

originating_object_id - This is the ID of the object that this knowledge chunk was derived from, such as an imported web-page or imported PDF document