Date management / latest information retrieval. Possible without GPT?

[raphael5]

Hi everyone,

Quick question here. I am dealing with a bunch of legal documents. As you may I have guessed, legal documents are sometimes dated, especially judgments / case law.

When retrieving such information with RAG, I have two challenges. Sometimes I want to organize such information in a chronological order. Sometimes I want to keep the most recent information as the oldest information would be irrelevant.

I haven’t found anything, but I was wondering if there was any RAG practice that could help handle such scenario?

If not, are there any alternatives? I guess I could do that using a prompt to GPT, but obviously this would increase the processing time & cost. Hence my question about cheaper methods.

Thanks a lot for your help

[PaulBellow]

This is a great thread about something @stevenic cooked up for RAG you might want to take a look at.

Document Sections: Better rendering of chunks for long documents

I wanted to share with the community some work I’ve been doing around Retrieval Augmented Generation (RAG) using long form documents. I’ve developed a new technique called “Document Sections” which I feel is superior to traditional techniques being employed today. Here’s an example of the chunks returned by a vector DB like Pinecone. This is the chunk output from Vectra, my local vector DB project, for a query over a small corpus of documents from the Teams AI Library I designed. [image] Th…