File size: 3,298 Bytes
b7424f7
d162d0d
b7424f7
 
 
 
 
 
 
 
 
 
789d89c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
---
title: CAROLL NER Demos
emoji: 🐠
colorFrom: purple
colorTo: pink
sdk: streamlit
sdk_version: 1.36.0
app_file: app.py
pinned: false
license: mit
---

#### German Legal NER:

This language model is trained on the [Legal Entity Recognition](https://github.com/elenanereiss/Legal-Entity-Recognition) dataset. We conducted a stratified 10-fold cross-validation to prevent overfitting. The results showed that their fine-tuned German BERT model outperformed the existing BiLSTM-CRF+ model, which was previously used on the same LER dataset. It is capable of annotating German legal data with the following 19 distinct labels:

|Abbreviation|Class|
|----|----|
|PER|Person|
|RR|Judge|
|AN|Lawyer|
|LD|Country|
|ST|City|
|STR|Street|
|LDS|Landscape|
|ORG|Organization|
|UN|Company|
|INN|Institution|
|GRT|Court|
|MRK|Brand|
|GS|Law|
|VO|Ordinance|
|EUN|European legal norm|
|VS|Regulation|
|VT|Contract|
|RS|Court decision|
|LIT|Legal literature|

This model is publicly available at [PaDaS-Lab/gbert-legal-ner](https://huggingface.co/PaDaS-Lab/gbert-legal-ner). We have also published a corresponding [paper](https://arxiv.org/pdf/2303.05388.pdf) in this regard. Please cite this paper while using this model:

```bibtex
@conference{icaart23,
  author={Harshil Darji. and Jelena Mitrović. and Michael Granitzer.},
  title={German BERT Model for Legal Named Entity Recognition},
  booktitle={Proceedings of the 15th International Conference on Agents and Artificial Intelligence - Volume 3: ICAART,},
  year={2023},
  pages={723-728},
  publisher={SciTePress},
  organization={INSTICC},
  doi={10.5220/0011749400003393},
  isbn={978-989-758-623-1},
  issn={2184-433X},
}
```
---
#### GDPR Privacy Policy NER:

This language model is trained on a privacy policy dataset. This dataset is annotated using 33 labels that are in accordance with GDPR. This model aims to facilitate information extraction related to GDPR from a given privacy policy. It can also be further improved to verify whether a given privacy policy follows the GDPR regulations. As stated above, this model is capable of annotating given privacy policy-related text with the following 33 labels:

|Abbreviation|Class|
|----|----|
|DC|Data Controller|
|DP|Data Processor|
|DPO|Data Protection Officer|
|R|Recipient|
|TP|Third Party|
|A|Authority|
|DS|Data Subject|
|DSO|Data Source|
|RP|Required Purpose|
|NRP|Not-Required Purpose|
|P|Processing|
|NPD|Non-Personal Data|
|PD|Personal Data|
|OM|Organisational Measure|
|TM|Technical Measure|
|LB|Legal Basis|
|CONS|Consent|
|CONT|Contract|
|LI|Legitimate Interest|
|ADM|Automated Decision Making|
|RET|Retention|
|SEU|Scale EU|
|SNEU|Scale Non-EU|
|RI|Right|
|DSR15|Art. 15 Right of access by the data subject|
|DSR16|Art. 16 Right to rectification|
|DSR17|Art. 17 Right to erasure ("right to be forgotten")|
|DSR18|Art. 18 Right to restriction of processing|
|DSR19|Art. 19 Notification obligation regarding rectification or erasure of personal data or restriction of processing|
|DSR20|Art. 20 Right to data portability|
|DSR21|Art. 21 Right to object|
|DSR22|Art. 22 Automated individual decision-making, including profiling|
|LC|Lodge Complaint|

This model is publicly available at [PaDaS-Lab/gdpr-privacy-policy-ner](https://huggingface.co/PaDaS-Lab/gdpr-privacy-policy-ner).