Skip to content

Custom named entity recognition (persons, locations) using spaCy for German texts annotated in Transkribus

License

Notifications You must be signed in to change notification settings

RISE-UNIBAS/transkribus-custom-ner-de

Repository files navigation

transkribus-custom-ner-de

Custom named entity recognition (persons, locations) using spaCy for German texts annotated in Transkribus.

Access this binder at the following URL

https://mybinder.org/v2/gh/RISE-UNIBAS/transkribus-custom-ner-de/HEAD?labpath=interactive.ipynb

DOI Binder

Creator

This software and sample dataset were created by the University of Basel's Research and Infrastructure Support RISE ([email protected]) in May 2022.

File structure and data overview

Tutorial

This tutorial aims to show you how to get from a document in Transkribus to a custom NER model.

First step: prepare your documents for training in Transkribus

This section explains how to prepare your documents using the Transkribus Expert Client in order to train a spaCy model for named entity recognition. Basic familiarity with the Transkribus Expert Client is assumed (see READ-COOP for tutorials). First the workflow is presented and then a detailed annotation sample is provided.

Workflow

The preparation workflow involves five sub-steps:

  1. Create a collection in Transkribus and upload your documents. Automatically or manually segment and transcribe the documents, and carry out corrections if necessary. Be sure that you are happy with the result since there is no going back.
  2. Choose a representative sample of all documents. Think about the required sample size and make sure to include as much variety as possible.
  3. Define which entities you want to annotate and specify annotation guidelines. Try to be explicit in recording all annotation decisions you are making.
  4. Annotate the representative sample. Try to ensure that annotations across all documents are consistent. Of course, annotation guidelines will likely have to be amended during the annotation process. Again, be sure that you are happy with the result since there is no going back.
  5. Export the annotated representative sample (again, there is no going back from here).

Sample annotation

For a general introduction on how to annotate text using the Transkribus Expert Client, see https://readcoop.eu/transkribus/howto/how-to-enrich-transcribed-documents-with-mark-up/.

Second step: train, evaluate, and apply a custom NER model using spaCy

This section explains how to train, evaluate, and apply a custom NER model using spaCy with the annotated representative sample created in the previous section. For this, either the Binder or the Client class of the transkribus_custom_ner_de module can be employed:

from __future__ import annotations
from transkribus_custom_ner_de.client import Client
import os.path

DIR = os.path.dirname(__file__)
PARENT_DIR = os.path.dirname(os.path.dirname(__file__))
WORD_REMOVE = ["Beifall", Händeklatschen"]
PERSON_NAMES = ['Gustav Gottheil']
LOCATION_NAMES = ['Boston', 'Roman']
GOLD_STANDARD = PARENT_DIR + "/sample/gold_standard.zip"
TEXT_PATH = PARENT_DIR + "/sample/text_unanalyzed.txt"

my_client = Client()
my_client.train_model(zip_url=GOLD_STANDARD,
                      word_remove=WORD_REMOVE,
                      person_names=PERSON_NAMES,
                      location_names=LOCATION_NAMES,
                      epochs=100,
                      _local=True)
my_client.save_model()
my_client.evaluate_model()
my_client.apply_model(text_url=TEXT_PATH,
                      _local=True)
my_client.result
my_client.save_result2csv()

License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.