HTR-United is a Github organization without any other form of legal personality. It aims at gathering HTR/OCR transcriptions of all periods and style of writing, mostly but not exclusively in French. It was born from the mere necessity -for projects- to possess potential ground truth to rapidly train models on smaller corpora.
Datasets shared or referenced with HTR-United must, at minimum, take the form of:
- an ensemble of ALTO XML and/or PAGE XML files containing either only informations on the segmentation, either the segmentation and the corresponding transcription;
- an ensemble of corresponding images. They can be shared in the form of a simple permalink to ressources hosted somewhere else, or can be the contact information necessary to request access to the images. It must be possible to recompose the link between the XML files and the image without any intermediary process such as changing the images' names;
- a documentation on the context in which the dataset was produced and the rules followed to segment and transcribed the documents. For Github repositories, this documentation is usually presented in the README.
A corpus can be sub-divided into smaller ensembles if it seems necessary.
If you need help to compose your repository, you can check our template!
Eventually, this organization will also aim at sharing -under free licenses- models suited for requested HTR processors. This will make it possible for projects with smaller capacities to benefit from ready-to-use models. Thus, if you share your data, and according to the rythm followed by the other members, you will soon be able to use such models.
However, keep in mind there exists a virtuous circle Transcription<->Training which will eventually, we hope, considerably improve the transcriptions created by young projects starting from scratch.
There are two cases:
- You already have data in a repository
- You don't have one and prefer to help the organization directly
It's rather convenient: you stay in control, and there's no issue with joining the organization. However, if you want your dataset to gain visibility, it seems important to us that you describe it here. Indeed, if you take benefit from data or models provided by HTR-United, you may as well contribute!
To do so, you just need to open an issue or request an update on the deposit repository by adding a YAML file generated with our form, presented as follows:
schema: https://htr-united.github.io/schema/2021-10-15/schema.json
title: My Example Dataset
url: https://link.to.repository
authors:
- name: John
surname: Doe
roles:
- transcriber
- name: Jeanne
surname: Dupont
roles:
- project-manager
description: A short description of the content of the dataset.
project-name: My Awesome Project
project-website: https://optional.link.to.project
language:
- fra
script:
- Latn
script-type: only-manuscript
time:
notBefore: '1830'
notAfter: '1875'
hands:
count: '1'
precision: exact
license:
- name: CC-BY 4.0
url: https://creativecommons.org/licenses/by/4.0/
format: Page-XML
volume:
- metric: pages
count: 42
- metric: lines
count: 420
- metric: characters
count: 4200
transcription-guidelines: A presentation of the rules established for the transcription.
Well, we'll be happy to get help from you. Open an issue here and we will gladly help to create and share your repository on HTR-United. Skills with git
are appreciated but, if you want to share data, we will help you. It's the purpose of this organization!
You can browse the content of the catalog from our website: here.
Here is an overview of the periods covered by the datasets documented in HTR-United's catalog!
To help you improve and guarantee the quality of your dataset, we developped a series of tools which can easily be add to your repositories. Check out our Tools webpage to see descriptions and demos!
-
(FR) Alix Chagué, Thibault Clérice, Laurent Romary. HTR-United : Mutualisons la vérité de terrain !. DHNord2021 - Publier, partager, réutiliser les données de la recherche : les data papers et leurs enjeux, MESHS, Nov 2021, Lille, France. ⟨hal-03398740⟩
-
(FR) Alix Chagué. Conditions de la mutualisation : les principes FAIR et HTR-United. Humanistica 2022, May 2022, Montréal, Canada. ⟨hal-03685731⟩
Logo by Alix Chagué.