FrancophonIA/GROBID-Dictionaries

该数据集是基于Auguste Laverdet发布的目录和Gabriel Charavay及Maison Charavay出版的拍卖销售目录的扫描。这些目录包括Gabriel Charavay的《Revue des autographes》第一系列和第二系列，Auguste Laverdet的《Catalogue de lettres autographes et manuscrits》，以及Etienne Charavay的拍卖目录。数据集包含四种类型的目录，每种类型都有15页经过注释的数据，其中10页用于训练，5页用于评估。数据通过Transkribus进行OCR处理，并按照GROBID字典推荐的方法进行标注。数据集的训练和评估数据可在`ToyData`文件夹中找到，遵循CC-BY许可。

The dataset is based on scans of catalogues published by Auguste Laverdet, as well as auction sales catalogues published by Gabriel Charavay and Maison Charavay. These include the first and second series of Gabriel Charavays Revue des autographes, Auguste Laverdets Catalogue de lettres autographes et manuscrits, and auction catalogues by Etienne Charavay. The dataset contains four types of catalogues, each with 15 annotated pages: 10 for training and 5 for evaluation. The data has been OCR-processed using Transkribus and annotated according to the recommendations of GROBID dictionaries. The training and evaluation data of the dataset can be found in the ToyData folder, and it follows the CC-BY license.