langgenius/dify
1
0
Fork 0
mirror of https://github.com/langgenius/dify.git synced 2026-07-21 12:05:30 -04:00
Code Issues 775 Packages Projects Releases 10 Wiki Activity

Improve Keyword Extraction Accuracy and Support External Parsing in RAG #16826

New Issue
Closed
opened 2026-02-21 19:27:59 -05:00 by yindo · 0 comments
yindo commented 2026-02-21 19:27:59 -05:00
Owner
Copy Link
Copy Source

Originally created by @louxingyu380 on GitHub (Sep 6, 2025).

Self Checks

  • I have read the Contributing Guide and Language Policy.
  • I have searched for existing issues search for existing issues, including closed ones.
  • I confirm that I am using English to submit this report, otherwise it will be closed.
  • Please do not modify this template :) and fill in all the required fields.

1. Is this request related to a challenge you're experiencing? Tell me about your story.

Hi Dify team, first of all, thank you for building such a powerful and flexible platform! We’ve been using Dify internally to build a corporate knowledge base assistant that helps employees quickly find policies, technical documentation, and project records.

At first, we were excited — the UI is clean, the workflow is intuitive, and connecting to our vector database was seamless. However, as we tested it in real scenarios, we hit a frustrating wall: our documents simply aren’t being recalled properly — even when the answers are right there in the text.

After digging deeper, we realized the root cause lies in two key areas:

Weak keyword extraction during document parsing
When we upload a document , the system extracts keywords like “the”, “is”, missing critical terms. As a result, queries fail to retrieve the right chunk — even though it exists.This severely limits the recall accuracy of the RAG pipeline.

No support for external document parsers
Our internal documents include scanned PDFs, forms, and engineering diagrams — formats that are hard to parse accurately with generic tools. We already have a high-accuracy parsing service in-house (using layout analysis and domain-specific models), but Dify doesn’t allow us to plug it in.
🛠️ If Dify could support custom parsing via external API calls, we could offload parsing to our own system and feed clean, structured text back into the pipeline.

✅ Suggested Improvements:
Improve keyword extraction using smarter methods
Allow users to configure an external parsing endpoint — e.g., send uploaded files to a user-defined API and use the returned text.
This would make Dify not just a great prototyping tool, but a production-ready platform for enterprise knowledge management.

Thanks again for your hard work — we’re excited to see where Dify goes next!

2. Additional context or comments

No response

3. Can you help us with this feature?

  • I am interested in contributing to this feature.
Originally created by @louxingyu380 on GitHub (Sep 6, 2025). ### Self Checks - [x] I have read the [Contributing Guide](https://github.com/langgenius/dify/blob/main/CONTRIBUTING.md) and [Language Policy](https://github.com/langgenius/dify/issues/1542). - [x] I have searched for existing issues [search for existing issues](https://github.com/langgenius/dify/issues), including closed ones. - [x] I confirm that I am using English to submit this report, otherwise it will be closed. - [x] Please do not modify this template :) and fill in all the required fields. ### 1. Is this request related to a challenge you're experiencing? Tell me about your story. Hi Dify team, first of all, thank you for building such a powerful and flexible platform! We’ve been using Dify internally to build a corporate knowledge base assistant that helps employees quickly find policies, technical documentation, and project records. At first, we were excited — the UI is clean, the workflow is intuitive, and connecting to our vector database was seamless. However, as we tested it in real scenarios, we hit a frustrating wall: our documents simply aren’t being recalled properly — even when the answers are right there in the text. After digging deeper, we realized the root cause lies in two key areas: Weak keyword extraction during document parsing When we upload a document , the system extracts keywords like “the”, “is”, missing critical terms. As a result, queries fail to retrieve the right chunk — even though it exists.This severely limits the recall accuracy of the RAG pipeline. No support for external document parsers Our internal documents include scanned PDFs, forms, and engineering diagrams — formats that are hard to parse accurately with generic tools. We already have a high-accuracy parsing service in-house (using layout analysis and domain-specific models), but Dify doesn’t allow us to plug it in. 🛠️ If Dify could support custom parsing via external API calls, we could offload parsing to our own system and feed clean, structured text back into the pipeline. ✅ Suggested Improvements: Improve keyword extraction using smarter methods Allow users to configure an external parsing endpoint — e.g., send uploaded files to a user-defined API and use the returned text. This would make Dify not just a great prototyping tool, but a production-ready platform for enterprise knowledge management. Thanks again for your hard work — we’re excited to see where Dify goes next! ### 2. Additional context or comments _No response_ ### 3. Can you help us with this feature? - [x] I am interested in contributing to this feature.
yindo added the 💪 enhancement👻 feat:rag labels 2026-02-21 19:27:59 -05:00
yindo closed this issue 2026-02-21 19:27:59 -05:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
Knowledge Pipeline
Proposal
WIP
cloud
dependencies
feat:webapp
good first issue
outdated
performance
pull-request
Mirrored from GitHub Pull Request
 python:uv
refactor
size:L
stale
status: accepting prs
status: blocked
status: in discussion
stt & tts
vulnerablity
web
wont-fix
🌊 feat:workflow
🌚 invalid
🐞 bug
👻 feat:rag
💪 enhancement
📚 documentation
🔨 feat:tools
🙋‍♂️ question
🙏 help wanted
🤔 cant-reproduce
🤖 feat:agent
No Label 👻 feat:rag 💪 enhancement
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: langgenius/dify#16826
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?