arc53/DocsGPT

Feature: Add graph retriever

Open

#1015 opened on Jun 27, 2024

View on GitHub
 (5 comments) (0 reactions) (1 assignee)Python (13,691 stars) (1,359 forks)batch import
applicationbackendhelp wantedpython

Description

I think we can make upload / training process more advanced, maybe let users choose some "advanced" options on upload in the UI.

This way we can for example prepare data for graph + vector database. And when user creates a query it will automatically figure out where which retriever type needs to be used for that source. So I guess its a two part issue:

  • Edit structure of source docs in general (clean up a bit, and add retriever type field)

  • Add new graph retriever, add graph ingest, add UI to choose ingestion method maybe

Contributor guide

Tech stack
pythonreactjavascript
Domain
full stackai
Issue type
feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
4
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-2 days
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
blocked
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
mostly clear
Prerequisites
PythonReactGraph databasesVector databases
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
20
Research direction
The issue proposes adding a graph retriever and advanced upload options. The existing codebase uses Python for backend and likely React for frontend. The contributor should examine the current document upload and retrieval flow in the 'api' and 'frontend' directories. There is an assignee, so coordinate with them first. Look into graph database integration (e.g., Neo4j) and how to extend the existing retriever classes. The UI changes involve adding a dropdown for ingestion method. Implementation should start with the backend changes to source document structure.