Coming soon - Get a detailed view of why an account is flagged as spam!
view details

This post has been de-listed

It is no longer included in search results and normal feeds (front page, hot posts, subreddit posts, etc). It remains visible only via the author's post history.

1
Chunk CSV Data to create a vectorstore
Post Flair (click to view more posts with a particular flair)
Post Body

Task: Query a CSV file (without using built-in agents)

Input: CSV file

Output: JSON object like{"column": , "value" , "row_ids":}

If I embed the data and use a retriever on the vectorestore using similarity_search, I do not get all the matching instances in my result (as I cannot just use a very large k value). I used the 'parser' approach and got decent results. Can anyone suggest a better approach to get more accurate results?

Thanks

Comments

https://preview.redd.it/9xvyz5f542yc1.png?width=3432&format=png&auto=webp&s=f79740c6b0d841fda2cb9cc42ec5e5d967cf1843

If I understand correctly, your problem is that your RAG retrieval system does not retrieve all the relevant context you want for a search query.

This has nothing to do with the parser, so I wouldn't ask for help with the parser, but rather I would ask for help building a better retrieval model. If I got this correct, then hopefully HyDE https://arxiv.org/abs/2212.10496 can help you out for your use case.

Author
Account Strength
80%
Account Age
4 years
Verified Email
Yes
Verified Flair
No
Total Karma
47
Link Karma
28
Comment Karma
19
Profile updated: 3 days ago

Subreddit

Post Details

We try to extract some basic information from the post title. This is not always successful or accurate, please use your best judgement and compare these values to the post title and body for confirmation.
Posted
6 months ago