Seed the rewrite: profile, calibration, and repo scaffolding
Ground-up replacement for news_triage, whose signal quality failed because it had no representation of what its user likes -- three keywords fed to BM25 plus a two-term substring blocklist. profile/ holds the taste model as prose and TOML: 11 interest blocks, 3 veto blocks, 5 importance blocks, each a set of probe sentences to be embedded. The [scoring] constants are fitted, not guessed. CALIBRATION.md records the measurements against all 3,885 items in the v0.2 database and 40 real feedback labels, including two findings that changed the design: - compare blocks by z-score within block, never raw cosine; raw cosine let geopolitics win 78% of items and buried known-liked items at median rank 1246 of 3885 (293 after normalizing) - veto needs a margin over the interest signal, not a threshold; a plain 0.55 cutoff falsely killed "Belgian mayor blocks club from hosting Israeli opponent in Europa League", which the old substring blocklist also killed
This commit is contained in:
commit
c89a723603
6 files changed
+551
No files matched your search
+32
@@ -0,0 +1,32 @@
|
||||
# Allowlist ignore file, per the house convention: ignore everything, then
|
||||
# explicitly re-admit tracked paths.
|
||||
**/*
|
||||
|
||||
!.gitignore
|
||||
!README.md
|
||||
!pyproject.toml
|
||||
!uv.lock
|
||||
|
||||
!profile/
|
||||
!profile/**
|
||||
|
||||
!news/
|
||||
!news/**
|
||||
|
||||
!tests/
|
||||
!tests/**
|
||||
|
||||
!deploy/
|
||||
!deploy/**
|
||||
|
||||
# Re-exclude generated and local artifacts that would otherwise slip through
|
||||
# the directory re-admissions above.
|
||||
news/**/__pycache__/**
|
||||
news/**/*.pyc
|
||||
tests/**/__pycache__/**
|
||||
tests/**/*.pyc
|
||||
**/*.db
|
||||
**/*.db-wal
|
||||
**/*.db-shm
|
||||
**/*.npy
|
||||
.venv/**
|
||||
Reference in new issue
Block a user