Re:Show HN: Danswer – Open-source question answering across ...
by PeterStuer on Monday July 10, 2023 @01:11PM
(#36669375)
In my experience the QA with documents pattern is fairly straightforward to implement. 90% of the effort to get to a preformant system hoever goes into massaging the documents into semantically meaningfull chuncks.
Most business documents, unlike blogposts and news articles, are not just running text. They have a lott of implicit structure that when lost as the typical naive chunckers do, lose much of the contextualized meaning as well.
[ Parent
| on HN ]
Re:Show HN: Danswer – Open-source question answering across ...
by Weves on Monday July 10, 2023 @05:40PM
(#36673336)
Agree with the point about intelligent chunking being very important! Each individual app connector can choose how it wants to split each `document` into `section`s (important point: this is customized at an app-level). The default chunker then keeps each section as part of a single chunk as much as possible. The goal here is, as you mentioned, to give each chunk the relevant surrounding context.
Additionally, the indexing process is setup as a composable pipeline under the hood. It would be fairly trivial to plug in different chunkers for different sources as needed in the future.
Re:Show HN: Danswer – Open-source question answering across ...
by darkteflon on Monday July 10, 2023 @07:34PM
(#36674522)
Chunking is very important but might, I feel, best be contextualised as one aspect of the bigger substantive challenge, which is how to prevent false negatives at the context retrieval stage - a.k.a. how to ensure your (vector? hybrid?) search returns all relevant context to the LLM’s context window.
Would you mind saying a few words on how Danswer approaches this?
Re:Show HN: Danswer – Open-source question answering across ...
by andy99 on Monday July 10, 2023 @07:17PM
(#36674330)
Yes agreed, tooling abounds, the work for anyone who's serious about this is customizing everything so it works with the idiosyncrasies of the documents and questions a customer has. I'm happy to talk to anyone who is interested, we are doing something like this for a company now.