elenanereiss commited on
Commit
9d7a396
1 Parent(s): 25ee654

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +42 -21
README.md CHANGED
@@ -69,6 +69,10 @@ train-eval-index:
69
 
70
  A dataset of Legal Documents from German federal court decisions for Named Entity Recognition. The dataset is human-annotated with 19 fine-grained entity classes. The dataset consists of approx. 67,000 sentences and contains 54,000 annotated entities. NER tags use the `BIO` tagging scheme.
71
 
 
 
 
 
72
  For more details see [https://arxiv.org/pdf/2003.13016v1.pdf](https://arxiv.org/pdf/2003.13016v1.pdf).
73
 
74
  ### Supported Tasks and Leaderboards
@@ -82,16 +86,17 @@ German
82
 
83
  ## Dataset Structure
84
  ### Data Instances
85
- ```
86
  {
87
  'id': '1',
88
  'tokens': ['Eine', 'solchermaßen', 'verzögerte', 'oder', 'bewusst', 'eingesetzte', 'Verkettung', 'sachgrundloser', 'Befristungen', 'schließt', '§', '14', 'Abs.', '2', 'Satz', '2', 'TzBfG', 'aus', '.'],
89
- 'ner_tags': [38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 3, 22, 22, 22, 22, 22, 22, 38, 38]
 
90
  }
91
  ```
92
  ### Data Fields
93
 
94
- ```
95
  {
96
  'id': Value(dtype='string', id=None),
97
  'tokens': Sequence(feature=Value(dtype='string', id=None),
@@ -138,7 +143,26 @@ German
138
  'O'],
139
  id=None),
140
  length=-1,
141
- id=None)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
142
  }
143
  ```
144
 
@@ -149,45 +173,43 @@ German
149
  | Input Sentences | 53384 | 6666 | 6673 |
150
 
151
 
152
- <!--
153
 
154
  ## Dataset Creation
155
 
156
  ### Curation Rationale
157
 
158
- [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
159
 
160
- -->
161
  ### Source Data
162
 
163
  Court decisions from 2017 and 2018 were selected for the dataset, published online by the [Federal Ministry of Justice and Consumer Protection](http://www.rechtsprechung-im-internet.de). The documents originate from seven federal courts: Federal Labour Court (BAG), Federal Fiscal Court (BFH), Federal Court of Justice (BGH), Federal Patent Court (BPatG), Federal Social Court (BSG), Federal Constitutional Court (BVerfG) and Federal Administrative Court (BVerwG).
164
 
165
- <!-- #### Initial Data Collection and Normalization
166
 
167
- [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
 
 
168
 
169
  #### Who are the source language producers?
170
 
171
- [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
172
- -->
173
 
174
  ### Annotations
175
 
176
- For more details see [https://github.com/elenanereiss/Legal-Entity-Recognition/blob/master/docs/Annotationsrichtlinien.pdf](https://github.com/elenanereiss/Legal-Entity-Recognition/blob/master/docs/Annotationsrichtlinien.pdf).
177
 
178
- <!-- #### Annotation process
179
 
180
- [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
181
 
182
- #### Who are the annotators?
183
-
184
- [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
185
 
186
  ### Personal and Sensitive Information
187
 
188
- [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
189
 
190
- ## Considerations for Using the Data
191
 
192
  ### Social Impact of Dataset
193
 
@@ -205,8 +227,7 @@ For more details see [https://github.com/elenanereiss/Legal-Entity-Recognition/b
205
 
206
  ### Dataset Curators
207
 
208
- [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
209
- -->
210
 
211
  ### Licensing Information
212
 
 
69
 
70
  A dataset of Legal Documents from German federal court decisions for Named Entity Recognition. The dataset is human-annotated with 19 fine-grained entity classes. The dataset consists of approx. 67,000 sentences and contains 54,000 annotated entities. NER tags use the `BIO` tagging scheme.
71
 
72
+ The dataset includes two different versions of annotations, one with a set of 19 fine-grained semantic classes (`ner_tags`) and another one with a set of 7 coarse-grained classes (`ner_coarse_tags`). There are 53,632 annotated entities in total, the majority of which (74.34 %) are legal entities, the others are person, location and organization (25.66 %).
73
+
74
+ ![](https://raw.githubusercontent.com/elenanereiss/Legal-Entity-Recognition/master/docs/Distribution.png)
75
+
76
  For more details see [https://arxiv.org/pdf/2003.13016v1.pdf](https://arxiv.org/pdf/2003.13016v1.pdf).
77
 
78
  ### Supported Tasks and Leaderboards
 
86
 
87
  ## Dataset Structure
88
  ### Data Instances
89
+ ```python
90
  {
91
  'id': '1',
92
  'tokens': ['Eine', 'solchermaßen', 'verzögerte', 'oder', 'bewusst', 'eingesetzte', 'Verkettung', 'sachgrundloser', 'Befristungen', 'schließt', '§', '14', 'Abs.', '2', 'Satz', '2', 'TzBfG', 'aus', '.'],
93
+ 'ner_tags': [38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 3, 22, 22, 22, 22, 22, 22, 38, 38],
94
+ 'ner_coarse_tags': [14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 2, 9, 9, 9, 9, 9, 9, 14, 14]
95
  }
96
  ```
97
  ### Data Fields
98
 
99
+ ```python
100
  {
101
  'id': Value(dtype='string', id=None),
102
  'tokens': Sequence(feature=Value(dtype='string', id=None),
 
143
  'O'],
144
  id=None),
145
  length=-1,
146
+ id=None),
147
+ 'ner_coarse_tags': Sequence(feature=ClassLabel(num_classes=15,
148
+ names=['B-LIT',
149
+ 'B-LOC',
150
+ 'B-NRM',
151
+ 'B-ORG',
152
+ 'B-PER',
153
+ 'B-REG',
154
+ 'B-RS',
155
+ 'I-LIT',
156
+ 'I-LOC',
157
+ 'I-NRM',
158
+ 'I-ORG',
159
+ 'I-PER',
160
+ 'I-REG',
161
+ 'I-RS',
162
+ 'O'],
163
+ id=None),
164
+ length=-1,
165
+ id=None)
166
  }
167
  ```
168
 
 
173
  | Input Sentences | 53384 | 6666 | 6673 |
174
 
175
 
176
+
177
 
178
  ## Dataset Creation
179
 
180
  ### Curation Rationale
181
 
182
+ Documents in the legal domain contain multiple references to named entities, especially domain-specific named entities, i. e., jurisdictions, legal institutions, etc. Legal documents are unique and differ greatly from newspaper texts. On the one hand, the occurrence of general-domain named entities is relatively rare. On the other hand, in concrete applications, crucial domain-specific entities need to be identified in a reliable way, such as designations of legal norms and references to other legal documents (laws, ordinances, regulations, decisions, etc.). Most NER solutions operate in the general or news domain, which makes them inapplicable to the analysis of legal documents. Accordingly, there is a great need for an NER-annotated dataset consisting of legal documents, including the corresponding development of a typology of semantic concepts and uniform annotation guidelines.
183
 
 
184
  ### Source Data
185
 
186
  Court decisions from 2017 and 2018 were selected for the dataset, published online by the [Federal Ministry of Justice and Consumer Protection](http://www.rechtsprechung-im-internet.de). The documents originate from seven federal courts: Federal Labour Court (BAG), Federal Fiscal Court (BFH), Federal Court of Justice (BGH), Federal Patent Court (BPatG), Federal Social Court (BSG), Federal Constitutional Court (BVerfG) and Federal Administrative Court (BVerwG).
187
 
188
+ #### Initial Data Collection and Normalization
189
 
190
+ From the table of [contents](http://www.rechtsprechung-im-internet.de/rii-toc.xml), 107 documents from each court were selected (see Table 1). The data was collected from the XML documents, i. e., it was extracted from the XML elements `Mitwirkung, Titelzeile, Leitsatz, Tenor, Tatbestand, Entscheidungsgründe, Gründen, abweichende Meinung, and sonstiger Titel`. The metadata at the beginning of the documents (name of court, date of decision, file number, European Case Law Identifier, document type, laws) and those that belonged to previous legal proceedings was deleted. Paragraph numbers were removed.
191
+
192
+ The extracted data was split into sentences, tokenised using [SoMaJo](https://github.com/tsproisl/SoMaJo) and manually annotated in [WebAnno](https://webanno.github.io/webanno/).
193
 
194
  #### Who are the source language producers?
195
 
196
+ The Federal Ministry of Justice and the Federal Office of Justice provide selected decisions. Court decisions were produced by humans.
 
197
 
198
  ### Annotations
199
 
200
+ #### Annotation process
201
 
202
+ For more details see [annotation guidelines](https://github.com/elenanereiss/Legal-Entity-Recognition/blob/master/docs/Annotationsrichtlinien.pdf) (in German).
203
 
204
+ <!-- #### Who are the annotators?
205
 
206
+ [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)-->
 
 
207
 
208
  ### Personal and Sensitive Information
209
 
210
+ A fundamental characteristic of the published decisions is that all personal information have been anonymised for privacy reasons. This affects the classes person, location and organization.
211
 
212
+ <!-- ## Considerations for Using the Data
213
 
214
  ### Social Impact of Dataset
215
 
 
227
 
228
  ### Dataset Curators
229
 
230
+ [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)-->
 
231
 
232
  ### Licensing Information
233