Dom/anonymisation
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Stop words : +15 FP supplémentaires (apyréxie, mode, retraitée, villes, ass, cat)

Browse Source 
Audit OGC 17/74 : ajout variante accentuée apyréxie, termes courants
(mode, retraitée, régression, tel) et noms de villes françaises pour
éviter leur masquage comme NOM_GLOBAL.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
Domi31tls
2026-02-27 16:46:04 +01:00
parent 7991436caa
commit 84be2a5176
1 changed files with 5 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
anonymizer_core_refactored_onnx.py
View File

@@ -367,7 +367,7 @@ _MEDICAL_STOP_WORDS_SET = {
# FP audit OGC 21 — termes médicaux/courants flagués NOM_GLOBAL # FP audit OGC 21 — termes médicaux/courants flagués NOM_GLOBAL
"alimentation", "augmentation", "amelioration", "amélioration", "alimentation", "augmentation", "amelioration", "amélioration",
"biliaire", "biliaires", "bili", "voies", "voie", "biliaire", "biliaires", "bili", "voies", "voie",
"apyrexie", "apyrétique", "apyretique", "apyrexie", "apyréxie", "apyrétique", "apyretique",
"clavulanique", "mecillinam", "sulfamides", "sulfamide", "clavulanique", "mecillinam", "sulfamides", "sulfamide",
"tazobactam", "temocilline", "ecoflac", "furanes", "furane", "tazobactam", "temocilline", "ecoflac", "furanes", "furane",
"exilar", "lipruzet", "mopral", "exilar", "lipruzet", "mopral",
@@ -382,7 +382,10 @@ _MEDICAL_STOP_WORDS_SET = {
"eupnéique", "eupneique", "normocarde", "normotendue", "eupnéique", "eupneique", "normocarde", "normotendue",
"variable", "dosage", "posologie", "variable", "dosage", "posologie",
# Abréviations diététiques/soins trackare # Abréviations diététiques/soins trackare
"bcy", "po2", "po1", "po3", "bha", "atg", "ras", "cat", "bcy", "po2", "po1", "po3", "bha", "atg", "ras", "cat", "ass",
# FP audit OGC 17 CRH
"mode", "retraitee", "retraité", "retraitée", "régression", "regression", "tel",
"strasbourg", "bordeaux", "toulouse", "paris", "lyon", "marseille", "bayonne",
# Spécialités/services récurrents comme FP NOM # Spécialités/services récurrents comme FP NOM
"cancérologie", "cancerologie", "réanimation", "reanimation", "cancérologie", "cancerologie", "réanimation", "reanimation",
"urologie", "néphrologie", "nephrologie", "hématologie", "hematologie", "urologie", "néphrologie", "nephrologie", "hématologie", "hematologie",