flemming-it
3e018f528a
Some checks failed
CI / Linux x86_64 (Forgejo) (push) Failing after 1m20s
NER anonymization of German legal texts (plain text + DOCX) via the JuraNER model by Harshil Darji (MIT). The transformer model runs in the judge-ner host service; this module forwards documents over loopback HTTP and maps responses onto the capability contract.
27 lines
1.2 KiB
YAML
27 lines
1.2 KiB
YAML
schema_version: 1
|
|
|
|
author:
|
|
name: "Flemming.AI"
|
|
email: chain@flemming.ai
|
|
url: https://flemming.ai
|
|
license: Apache-2.0
|
|
|
|
tagline:
|
|
en: "NER anonymization for German legal texts (JuraNER)"
|
|
de: "NER-Anonymisierung für deutsche Rechtstexte (JuraNER)"
|
|
|
|
description:
|
|
en: "Transformer-based anonymization of German legal documents (plain text and DOCX) using the JuraNER model by Harshil Darji (MIT, HTW Berlin research). Detects persons, addresses, courts, file numbers and more; replaces them with consistent ⟦TYPE#⟧ tokens. Requires the judge-ner host service (Docker) — the NER model runs outside the WASM sandbox."
|
|
de: "Transformer-basierte Anonymisierung deutscher Rechtsdokumente (Klartext und DOCX) mit dem JuraNER-Modell von Harshil Darji (MIT, HTW-Berlin-Forschung). Erkennt Personen, Adressen, Gerichte, Aktenzeichen u. v. m. und ersetzt sie durch konsistente ⟦TYP#⟧-Token. Benötigt den judge-ner-Host-Service (Docker) — das NER-Modell läuft außerhalb der WASM-Sandbox."
|
|
|
|
category: text-processing
|
|
tags: [legal, german, anonymization, ner, gdpr, requires-host-service]
|
|
|
|
assets:
|
|
icon: assets/icon.svg
|
|
|
|
documentation:
|
|
long_description: MODULE.md
|
|
|
|
links:
|
|
repository: https://git.flemming.ai/chain-modules/text-anonymize-legal-de
|