Skip to content

docs: mark the gap-baseline entity-resolution entry resolved #442

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
seonghobae wants to merge 2 commits into from
Closed

docs: mark the gap-baseline entity-resolution entry resolved #442

Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
aa423bf
docs: mark the gap-baseline entity-resolution entry resolved
seonghobae Aug 22, 2026
8e6f24d
fix(frontend): use OIDC return-url helpers on the login button
seonghobae Aug 22, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion docs/product-technical-gap-baseline.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,7 @@
- **Multiple Project Extraction**: (Resolved) LLM prompt updated to request key_events as objects with project_name, separating events correctly.
- **5W1H Missing**: (Resolved) LLM prompt updated to explicitly request 5W1H evidence items in the JSON output array.
- **R&R and Keyman Missing**: (Resolved) LLM prompt updated to explicitly instruct using actual stated names rather than collective titles.
- **Entity Resolution / Searxng**: Abbreviations like "한전" and "한국전력" are not mapped properly using Searxng and KG corroboration.
- **Entity Resolution / Searxng**: (Resolved) `lineageweave/organization_name_resolution.py` (ADR 0008, SKOS `prefLabel`/`altLabel`-grounded) proposes a full canonical name for an abbreviated/slang organization mention via LLM, then cross-verifies it against `lineageweave/relation_verification.py`'s `SearxngRelationVerificationClient` before it is trusted -- reusing that module's Searxng client rather than a second web-search integration. Wired into the live keyman-extraction path (`_organization_name_resolution_client()` / `_relation_verification_client()` passed to `ingest_post_keymen` in `backend/app/main.py`'s `POST /api/posts/{post_id}/extract-keymen`), not just implemented and unused. The resolution prompt explicitly names "Korean-style contraction" as a case it handles (e.g. an "AGP"/"Aurora Grid Power"-style abbreviation), so "한전" → "한국전력" is the shape of mention this pipeline is built for.
- **Meso-level Team Mapping**: (Resolved) Checked extraction logic; `team` mapping logic is present and correct, but LLM needed better explicit instruction which is covered by R&R resolution.
- **Base64 Image Omni-modal**: Current text-only embedding fails on images. Omni-modal LLM processing is required for images to capture layout, font size, colors, and spatial meaning.

Expand Down
Loading