TimKoornstra commited on
Commit
a164599
·
verified ·
1 Parent(s): b90213a

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +132 -3
README.md CHANGED
@@ -1,3 +1,132 @@
1
- ---
2
- license: cc-by-4.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-4.0
3
+ task_categories:
4
+ - token-classification
5
+ language:
6
+ - nl
7
+ tags:
8
+ - ner
9
+ - named-entity-extraction
10
+ - dutch
11
+ - historical-documents
12
+ - archival-texts
13
+ - relabeling
14
+ - datasets
15
+ - 17th-century
16
+ - 18th-century
17
+ - 19th-century
18
+ pretty_name: Dutch Historical Notarial NER Dataset (Tag de Tekst)
19
+ ---
20
+
21
+ # Dutch Historical Notarial NER Dataset
22
+
23
+ This dataset is a relabeled version of the "AI-trainingset voor Named Entity Recognition (NER)" created during the crowdsourcing project [**"Tag de Tekst"** on VeleHanden.nl](https://taalmaterialen.ivdnt.org/download/aitrainingset1-0/) in 2020. It has been adapted for use in Named Entity Recognition (NER) tasks, with relabeling conducted using **[Google Deepmind's Gemini 2.0 Flash](https://ai.google.dev/gemini-api/docs/models/gemini-v2)** model.
24
+
25
+ ## Dataset Overview
26
+
27
+ - **Original Source**: Transcriptions of Dutch notarial texts from the 17th to 19th centuries.
28
+ - **Annotations**: Annotated by ~150 volunteers and reviewed by super users.
29
+ - **Relabeling**: Automatic relabeling into 4 entity classes:
30
+ - `persoon` (person names)
31
+ - `locatie` (locations)
32
+ - `datum` (dates)
33
+ - `organisatie` (organizations)
34
+ - **Sources**: Includes material from:
35
+ - Stadsarchief Amsterdam
36
+ - Nationaal Archief
37
+ - Noord-Hollands Archief
38
+ - Other regional historical centers
39
+ - **Total Scans**: 10,567
40
+ - **Language**: Dutch
41
+
42
+ ## Format
43
+
44
+ This dataset is formatted for use with [GLiNER](https://github.com/urchade/GLiNER). Each sample includes:
45
+ - `text`: The full text.
46
+ - `tokenized_text`: The text split into full-word tokens.
47
+ - `ner`: A list of annotated entities with start and end token indices and entity types.
48
+
49
+ Example:
50
+ ```json
51
+ {
52
+ "text": "Henrick Cardamon Op huijden ...",
53
+ "tokenized_text": ["Henrick", "Cardamon", "Op", "huijden", ...],
54
+ "ner": [
55
+ {
56
+ "start": 0,
57
+ "end": 1,
58
+ "label": "persoon"
59
+ },
60
+ ...
61
+ ]
62
+ }
63
+ ```
64
+
65
+ ## Usage
66
+
67
+ To load this dataset with the [Hugging Face Datasets](https://huggingface.co/docs/datasets/index) library, run:
68
+
69
+ ```py
70
+ from datasets import load_dataset
71
+
72
+ dataset = load_dataset("TimKoornstra/dutch-notarial-ner")
73
+ ```
74
+
75
+ ## Data Preprocessing and Relabeling
76
+
77
+ The original dataset, created during the **"Tag de Tekst"** project, was annotated by a large group of volunteers. While this collaborative effort provided a valuable starting point, the annotations were often inconsistent and contained inaccuracies. Common issues included:
78
+
79
+ - Mislabeling of entities (e.g., locations marked as persons).
80
+ - Overlapping or incomplete entity spans.
81
+ - Inconsistent application of annotation guidelines.
82
+
83
+ To address these challenges, the dataset was automatically relabeled using **[Google Deepmind's Gemini 2.0 Flash](https://ai.google.dev/gemini-api/docs/models/gemini-v2)** model. This process mapped all annotations into a simplified schema with four entity types:
84
+ - `persoon` (person names),
85
+ - `locatie` (locations),
86
+ - `datum` (dates),
87
+ - `organisatie` (organizations).
88
+
89
+ Additionally:
90
+ - **Inconsistent spans** were corrected to ensure uniformity.
91
+ - The data was reformatted for compatibility with modern tools like [GLiNER](https://github.com/urchade/GLiNER) and the Hugging Face `datasets` library.
92
+
93
+ These preprocessing steps ensure that the dataset is more accurate and consistent for training and evaluating Named Entity Recognition (NER) models.
94
+
95
+ ## Limitations
96
+
97
+ Despite preprocessing and relabeling, the dataset has some limitations:
98
+
99
+ - **Incomplete Entity Coverage**: While many errors were corrected, there may still be missed entities or incorrect spans, especially in complex cases.
100
+ - **Model-Induced Bias**: The relabeling process relied on **[Google Deepmind's Gemini 2.0 Flash](https://ai.google.dev/gemini-api/docs/models/gemini-v2)** model, which may introduce biases inherent to the model's training data.
101
+ - **Historical Context Challenges**: The dataset consists of historical Dutch texts (17th–19th century) with archaic language and formatting, which may pose additional challenges for modern models.
102
+ - **Potential Noise**: Due to the automatic relabeling process, there may still be minor inconsistencies or errors in the annotations.
103
+ - **HTR Artifacts**: The dataset is based on handwritten text recognition (HTR) outputs, so any transcription errors from the HTR process remain in the data. This limitation is consistent with the original dataset.
104
+
105
+ Users of this dataset should carefully evaluate its performance on their specific use case and consider further fine-tuning or validation if needed.
106
+
107
+ ## License
108
+
109
+ This dataset respects the original license: [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).
110
+
111
+ ## Citation
112
+
113
+ If you use this dataset in your research, please cite the original dataset and this repository:
114
+
115
+ ```bibtex
116
+ @misc{dutch_notarial_ner,
117
+ author = {Tim Koornstra},
118
+ title = {Dutch Historical Notarial NER Dataset},
119
+ year = {2025},
120
+ howpublished = {\url{https://huggingface.co/TimKoornstra/dutch-notarial-ner}},
121
+ note = {Relabeled with Gemini 2.0 Flash model}
122
+ }
123
+ ```
124
+
125
+ ## Acknowledgements
126
+
127
+ This dataset was originally developed as part of the projects:
128
+
129
+ - "De IJsberg zichtbaar maken" ([zoekintranscripties.nl](https://www.zoekintranscripties.nl/))
130
+ - "Slimmer zoeken in archieven" ([archieveninbeeld.nl](https://archieveninbeeld.nl/))
131
+
132
+ Special thanks to the volunteers of the "Tag de Tekst" project and the organizations contributing archival material.