Datasets:
Tasks:
Token Classification
Sub-tasks:
named-entity-recognition
Languages:
Portuguese
Size:
n<1K
License:
Commit
•
fcf3ca7
0
Parent(s):
Update files from the datasets library (from 1.2.0)
Browse filesRelease notes: https://github.com/huggingface/datasets/releases/tag/1.2.0
- .gitattributes +27 -0
- README.md +174 -0
- dataset_infos.json +1 -0
- dummy/default/1.0.0/dummy_data.zip +3 -0
- dummy/selective/1.0.0/dummy_data.zip +3 -0
- harem.py +311 -0
.gitattributes
ADDED
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
*.7z filter=lfs diff=lfs merge=lfs -text
|
2 |
+
*.arrow filter=lfs diff=lfs merge=lfs -text
|
3 |
+
*.bin filter=lfs diff=lfs merge=lfs -text
|
4 |
+
*.bin.* filter=lfs diff=lfs merge=lfs -text
|
5 |
+
*.bz2 filter=lfs diff=lfs merge=lfs -text
|
6 |
+
*.ftz filter=lfs diff=lfs merge=lfs -text
|
7 |
+
*.gz filter=lfs diff=lfs merge=lfs -text
|
8 |
+
*.h5 filter=lfs diff=lfs merge=lfs -text
|
9 |
+
*.joblib filter=lfs diff=lfs merge=lfs -text
|
10 |
+
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
11 |
+
*.model filter=lfs diff=lfs merge=lfs -text
|
12 |
+
*.msgpack filter=lfs diff=lfs merge=lfs -text
|
13 |
+
*.onnx filter=lfs diff=lfs merge=lfs -text
|
14 |
+
*.ot filter=lfs diff=lfs merge=lfs -text
|
15 |
+
*.parquet filter=lfs diff=lfs merge=lfs -text
|
16 |
+
*.pb filter=lfs diff=lfs merge=lfs -text
|
17 |
+
*.pt filter=lfs diff=lfs merge=lfs -text
|
18 |
+
*.pth filter=lfs diff=lfs merge=lfs -text
|
19 |
+
*.rar filter=lfs diff=lfs merge=lfs -text
|
20 |
+
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
21 |
+
*.tar.* filter=lfs diff=lfs merge=lfs -text
|
22 |
+
*.tflite filter=lfs diff=lfs merge=lfs -text
|
23 |
+
*.tgz filter=lfs diff=lfs merge=lfs -text
|
24 |
+
*.xz filter=lfs diff=lfs merge=lfs -text
|
25 |
+
*.zip filter=lfs diff=lfs merge=lfs -text
|
26 |
+
*.zstandard filter=lfs diff=lfs merge=lfs -text
|
27 |
+
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
@@ -0,0 +1,174 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
annotations_creators:
|
3 |
+
- expert-generated
|
4 |
+
language_creators:
|
5 |
+
- found
|
6 |
+
languages:
|
7 |
+
- pt
|
8 |
+
licenses:
|
9 |
+
- unknown
|
10 |
+
multilinguality:
|
11 |
+
- monolingual
|
12 |
+
size_categories:
|
13 |
+
- n<1K
|
14 |
+
source_datasets:
|
15 |
+
- original
|
16 |
+
task_categories:
|
17 |
+
- structure-prediction
|
18 |
+
task_ids:
|
19 |
+
- named-entity-recognition
|
20 |
+
---
|
21 |
+
|
22 |
+
# Dataset Card for [Dataset Name]
|
23 |
+
|
24 |
+
## Table of Contents
|
25 |
+
- [Dataset Description](#dataset-description)
|
26 |
+
- [Dataset Summary](#dataset-summary)
|
27 |
+
- [Supported Tasks](#supported-tasks-and-leaderboards)
|
28 |
+
- [Languages](#languages)
|
29 |
+
- [Dataset Structure](#dataset-structure)
|
30 |
+
- [Data Instances](#data-instances)
|
31 |
+
- [Data Fields](#data-fields)
|
32 |
+
- [Data Splits](#data-splits)
|
33 |
+
- [Dataset Creation](#dataset-creation)
|
34 |
+
- [Curation Rationale](#curation-rationale)
|
35 |
+
- [Source Data](#source-data)
|
36 |
+
- [Annotations](#annotations)
|
37 |
+
- [Personal and Sensitive Information](#personal-and-sensitive-information)
|
38 |
+
- [Considerations for Using the Data](#considerations-for-using-the-data)
|
39 |
+
- [Social Impact of Dataset](#social-impact-of-dataset)
|
40 |
+
- [Discussion of Biases](#discussion-of-biases)
|
41 |
+
- [Other Known Limitations](#other-known-limitations)
|
42 |
+
- [Additional Information](#additional-information)
|
43 |
+
- [Dataset Curators](#dataset-curators)
|
44 |
+
- [Licensing Information](#licensing-information)
|
45 |
+
- [Citation Information](#citation-information)
|
46 |
+
|
47 |
+
## Dataset Description
|
48 |
+
|
49 |
+
- **Homepage:** [HAREM homepage](https://www.linguateca.pt/primeiroHAREM/harem_coleccaodourada_en.html)
|
50 |
+
- **Repository:** [HAREM repository](https://www.linguateca.pt/primeiroHAREM/harem_coleccaodourada_en.html)
|
51 |
+
- **Paper:** [HAREM: An Advanced NER Evaluation Contest for Portuguese](http://comum.rcaap.pt/bitstream/10400.26/76/1/SantosSecoCardosoVilelaLREC2006.pdf)
|
52 |
+
- **Point of Contact:** [Diana Santos](mailto:diana.santos@sintef.no)
|
53 |
+
|
54 |
+
### Dataset Summary
|
55 |
+
|
56 |
+
The HAREM is a Portuguese language corpus commonly used for Named Entity Recognition tasks. It includes about 93k words, from 129 different texts,
|
57 |
+
from several genres, and language varieties. The split of this dataset version follows the division made by [1], where 7% HAREM
|
58 |
+
documents are the validation set and the miniHAREM corpus (with about 65k words) is the test set. There are two versions of the dataset set,
|
59 |
+
a version that has a total of 10 different named entity classes (Person, Organization, Location, Value, Date, Title, Thing, Event,
|
60 |
+
Abstraction, and Other) and a "selective" version with only 5 classes (Person, Organization, Location, Value, and Date).
|
61 |
+
|
62 |
+
It's important to note that the original version of the HAREM dataset has 2 levels of NER details, namely "Category" and "Sub-type".
|
63 |
+
The dataset version processed here ONLY USE the "Category" level of the original dataset.
|
64 |
+
|
65 |
+
[1] Souza, Fábio, Rodrigo Nogueira, and Roberto Lotufo. "BERTimbau: Pretrained BERT Models for Brazilian Portuguese." Brazilian Conference on Intelligent Systems. Springer, Cham, 2020.
|
66 |
+
|
67 |
+
### Supported Tasks and Leaderboards
|
68 |
+
|
69 |
+
[More Information Needed]
|
70 |
+
|
71 |
+
### Languages
|
72 |
+
|
73 |
+
Portuguese
|
74 |
+
|
75 |
+
## Dataset Structure
|
76 |
+
|
77 |
+
### Data Instances
|
78 |
+
|
79 |
+
```
|
80 |
+
{
|
81 |
+
"id": "HAREM-871-07800",
|
82 |
+
"ner_tags": [3, 0, 0, 3, 4, 4, 4, 4, 4, 4, 4, 4,
|
83 |
+
],
|
84 |
+
"tokens": [
|
85 |
+
"Abraço", "Página", "Principal", "ASSOCIAÇÃO", "DE", "APOIO", "A", "PESSOAS", "COM", "VIH", "/", "SIDA"
|
86 |
+
]
|
87 |
+
}
|
88 |
+
```
|
89 |
+
|
90 |
+
### Data Fields
|
91 |
+
|
92 |
+
- `id`: id of the sample
|
93 |
+
- `tokens`: the tokens of the example text
|
94 |
+
- `ner_tags`: the NER tags of each token
|
95 |
+
|
96 |
+
The NER tags correspond to this list:
|
97 |
+
```
|
98 |
+
"O", "B-PESSOA", "I-PESSOA", "B-ORGANIZACAO", "I-ORGANIZACAO", "B-LOCAL", "I-LOCAL", "B-TEMPO", "I-TEMPO", "B-VALOR", "I-VALOR", "B-ABSTRACCAO", "I-ABSTRACCAO", "B-ACONTECIMENTO", "I-ACONTECIMENTO", "B-COISA", "I-COISA", "B-OBRA", "I-OBRA", "B-OUTRO", "I-OUTRO"
|
99 |
+
```
|
100 |
+
|
101 |
+
The NER tags have the same format as in the CoNLL shared task: a B denotes the first item of a phrase and an I any non-initial word.
|
102 |
+
|
103 |
+
### Data Splits
|
104 |
+
|
105 |
+
The data is split into train, validation and test set for each of the two versions (default and selective). The split sizes are as follow:
|
106 |
+
|
107 |
+
| Train | Val | Test |
|
108 |
+
| ------ | ----- | ---- |
|
109 |
+
| 121 | 8 | 128 |
|
110 |
+
|
111 |
+
## Dataset Creation
|
112 |
+
|
113 |
+
### Curation Rationale
|
114 |
+
|
115 |
+
[More Information Needed]
|
116 |
+
|
117 |
+
### Source Data
|
118 |
+
|
119 |
+
#### Initial Data Collection and Normalization
|
120 |
+
|
121 |
+
[More Information Needed]
|
122 |
+
|
123 |
+
#### Who are the source language producers?
|
124 |
+
|
125 |
+
[More Information Needed]
|
126 |
+
|
127 |
+
### Annotations
|
128 |
+
|
129 |
+
#### Annotation process
|
130 |
+
|
131 |
+
[More Information Needed]
|
132 |
+
|
133 |
+
#### Who are the annotators?
|
134 |
+
|
135 |
+
[More Information Needed]
|
136 |
+
|
137 |
+
### Personal and Sensitive Information
|
138 |
+
|
139 |
+
[More Information Needed]
|
140 |
+
|
141 |
+
## Considerations for Using the Data
|
142 |
+
|
143 |
+
### Social Impact of Dataset
|
144 |
+
|
145 |
+
[More Information Needed]
|
146 |
+
|
147 |
+
### Discussion of Biases
|
148 |
+
|
149 |
+
[More Information Needed]
|
150 |
+
|
151 |
+
### Other Known Limitations
|
152 |
+
|
153 |
+
[More Information Needed]
|
154 |
+
|
155 |
+
## Additional Information
|
156 |
+
|
157 |
+
### Dataset Curators
|
158 |
+
|
159 |
+
[More Information Needed]
|
160 |
+
|
161 |
+
### Licensing Information
|
162 |
+
|
163 |
+
[More Information Needed]
|
164 |
+
|
165 |
+
### Citation Information
|
166 |
+
|
167 |
+
```
|
168 |
+
@inproceedings{santos2006harem,
|
169 |
+
title={Harem: An advanced ner evaluation contest for portuguese},
|
170 |
+
author={Santos, Diana and Seco, Nuno and Cardoso, Nuno and Vilela, Rui},
|
171 |
+
booktitle={quot; In Nicoletta Calzolari; Khalid Choukri; Aldo Gangemi; Bente Maegaard; Joseph Mariani; Jan Odjik; Daniel Tapias (ed) Proceedings of the 5 th International Conference on Language Resources and Evaluation (LREC'2006)(Genoa Italy 22-28 May 2006)},
|
172 |
+
year={2006}
|
173 |
+
}
|
174 |
+
```
|
dataset_infos.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"default": {"description": "\nThe HAREM is a Portuguese language corpus commonly used for Named Entity Recognition tasks. It includes about 93k words, from 129 different texts,\nfrom several genres, and language varieties. The split of this dataset version follows the division made by [1], where 7% HAREM\ndocuments are the validation set and the miniHAREM corpus (with about 65k words) is the test set. There are two versions of the dataset set,\na version that has a total of 10 different named entity classes (Person, Organization, Location, Value, Date, Title, Thing, Event,\nAbstraction, and Other) and a \"selective\" version with only 5 classes (Person, Organization, Location, Value, and Date).\n\nIt's important to note that the original version of the HAREM dataset has 2 levels of NER details, namely \"Category\" and \"Sub-type\".\nThe dataset version processed here ONLY USE the \"Category\" level of the original dataset.\n\n[1] Souza, F\u00e1bio, Rodrigo Nogueira, and Roberto Lotufo. \"BERTimbau: Pretrained BERT Models for Brazilian Portuguese.\" Brazilian Conference on Intelligent Systems. Springer, Cham, 2020.\n", "citation": "\n@inproceedings{santos2006harem,\n title={Harem: An advanced ner evaluation contest for portuguese},\n author={Santos, Diana and Seco, Nuno and Cardoso, Nuno and Vilela, Rui},\n booktitle={quot; In Nicoletta Calzolari; Khalid Choukri; Aldo Gangemi; Bente Maegaard; Joseph Mariani; Jan Odjik; Daniel Tapias (ed) Proceedings of the 5 th International Conference on Language Resources and Evaluation (LREC'2006)(Genoa Italy 22-28 May 2006)},\n year={2006}\n}\n", "homepage": "https://www.linguateca.pt/primeiroHAREM/harem_coleccaodourada_en.html", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "tokens": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "ner_tags": {"feature": {"num_classes": 21, "names": ["O", "B-PESSOA", "I-PESSOA", "B-ORGANIZACAO", "I-ORGANIZACAO", "B-LOCAL", "I-LOCAL", "B-TEMPO", "I-TEMPO", "B-VALOR", "I-VALOR", "B-ABSTRACCAO", "I-ABSTRACCAO", "B-ACONTECIMENTO", "I-ACONTECIMENTO", "B-COISA", "I-COISA", "B-OBRA", "I-OBRA", "B-OUTRO", "I-OUTRO"], "names_file": null, "id": null, "_type": "ClassLabel"}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "builder_name": "harem", "config_name": "default", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1506373, "num_examples": 121, "dataset_name": "harem"}, "test": {"name": "test", "num_bytes": 1062714, "num_examples": 128, "dataset_name": "harem"}, "validation": {"name": "validation", "num_bytes": 51318, "num_examples": 8, "dataset_name": "harem"}}, "download_checksums": {"https://raw.githubusercontent.com/neuralmind-ai/portuguese-bert/master/ner_evaluation/data/FirstHAREM-total-train.json": {"num_bytes": 1060674, "checksum": "3542944e1e56145c5d1f4df1750df8ec81d0f9e0a7cc0c3e74b0b26df5869763"}, "https://raw.githubusercontent.com/neuralmind-ai/portuguese-bert/master/ner_evaluation/data/FirstHAREM-total-dev.json": {"num_bytes": 47603, "checksum": "a08705c45caef5bdb82b7fe394491de114edb24d14e007a9f3978be030219537"}, "https://raw.githubusercontent.com/neuralmind-ai/portuguese-bert/master/ner_evaluation/data/MiniHAREM-total.json": {"num_bytes": 779004, "checksum": "9a31f28df9664d7de4ceab6f2ec427ad1761463348083d35c7fa97cae87505db"}}, "download_size": 1887281, "post_processing_size": null, "dataset_size": 2620405, "size_in_bytes": 4507686}, "selective": {"description": "\nThe HAREM is a Portuguese language corpus commonly used for Named Entity Recognition tasks. It includes about 93k words, from 129 different texts,\nfrom several genres, and language varieties. The split of this dataset version follows the division made by [1], where 7% HAREM\ndocuments are the validation set and the miniHAREM corpus (with about 65k words) is the test set. There are two versions of the dataset set,\na version that has a total of 10 different named entity classes (Person, Organization, Location, Value, Date, Title, Thing, Event,\nAbstraction, and Other) and a \"selective\" version with only 5 classes (Person, Organization, Location, Value, and Date).\n\nIt's important to note that the original version of the HAREM dataset has 2 levels of NER details, namely \"Category\" and \"Sub-type\".\nThe dataset version processed here ONLY USE the \"Category\" level of the original dataset.\n\n[1] Souza, F\u00e1bio, Rodrigo Nogueira, and Roberto Lotufo. \"BERTimbau: Pretrained BERT Models for Brazilian Portuguese.\" Brazilian Conference on Intelligent Systems. Springer, Cham, 2020.\n", "citation": "\n@inproceedings{santos2006harem,\n title={Harem: An advanced ner evaluation contest for portuguese},\n author={Santos, Diana and Seco, Nuno and Cardoso, Nuno and Vilela, Rui},\n booktitle={quot; In Nicoletta Calzolari; Khalid Choukri; Aldo Gangemi; Bente Maegaard; Joseph Mariani; Jan Odjik; Daniel Tapias (ed) Proceedings of the 5 th International Conference on Language Resources and Evaluation (LREC'2006)(Genoa Italy 22-28 May 2006)},\n year={2006}\n}\n", "homepage": "https://www.linguateca.pt/primeiroHAREM/harem_coleccaodourada_en.html", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "tokens": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "ner_tags": {"feature": {"num_classes": 11, "names": ["O", "B-PESSOA", "I-PESSOA", "B-ORGANIZACAO", "I-ORGANIZACAO", "B-LOCAL", "I-LOCAL", "B-TEMPO", "I-TEMPO", "B-VALOR", "I-VALOR"], "names_file": null, "id": null, "_type": "ClassLabel"}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "builder_name": "harem", "config_name": "selective", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1506373, "num_examples": 121, "dataset_name": "harem"}, "test": {"name": "test", "num_bytes": 1062714, "num_examples": 128, "dataset_name": "harem"}, "validation": {"name": "validation", "num_bytes": 51318, "num_examples": 8, "dataset_name": "harem"}}, "download_checksums": {"https://raw.githubusercontent.com/neuralmind-ai/portuguese-bert/master/ner_evaluation/data/FirstHAREM-selective-train.json": {"num_bytes": 969734, "checksum": "afb49c5d11116ff297d7abb7657f524917cb5704b221d5e3fb687e064a71e494"}, "https://raw.githubusercontent.com/neuralmind-ai/portuguese-bert/master/ner_evaluation/data/FirstHAREM-selective-dev.json": {"num_bytes": 38988, "checksum": "2ea2d350c587d35b08a86d067f6de27df6e3587339e80d10df787c7443fca7f3"}, "https://raw.githubusercontent.com/neuralmind-ai/portuguese-bert/master/ner_evaluation/data/MiniHAREM-selective.json": {"num_bytes": 707151, "checksum": "7a5d88cf1319ddae1940a02d3fde7dd5841863e05e616cfb2b574613407b7f37"}}, "download_size": 1715873, "post_processing_size": null, "dataset_size": 2620405, "size_in_bytes": 4336278}}
|
dummy/default/1.0.0/dummy_data.zip
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c24820cb6b58636abc0c1275fb1b02c4897b6cb6ef0bfbf5f99f31a06be5d87f
|
3 |
+
size 29509
|
dummy/selective/1.0.0/dummy_data.zip
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:683f2e657970d34b20dc17fc756474f75ffbffb5d5bd8c2b632ada16312c4315
|
3 |
+
size 27902
|
harem.py
ADDED
@@ -0,0 +1,311 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# coding=utf-8
|
2 |
+
# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
|
3 |
+
#
|
4 |
+
# Licensed under the Apache License, Version 2.0 (the "License");
|
5 |
+
# you may not use this file except in compliance with the License.
|
6 |
+
# You may obtain a copy of the License at
|
7 |
+
#
|
8 |
+
# http://www.apache.org/licenses/LICENSE-2.0
|
9 |
+
#
|
10 |
+
# Unless required by applicable law or agreed to in writing, software
|
11 |
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
12 |
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
13 |
+
# See the License for the specific language governing permissions and
|
14 |
+
# limitations under the License.
|
15 |
+
"""HAREM dataset"""
|
16 |
+
|
17 |
+
from __future__ import absolute_import, division, print_function
|
18 |
+
|
19 |
+
import json
|
20 |
+
import logging
|
21 |
+
import unicodedata
|
22 |
+
from typing import List, Tuple
|
23 |
+
|
24 |
+
import datasets
|
25 |
+
|
26 |
+
|
27 |
+
_CITATION = """
|
28 |
+
@inproceedings{santos2006harem,
|
29 |
+
title={Harem: An advanced ner evaluation contest for portuguese},
|
30 |
+
author={Santos, Diana and Seco, Nuno and Cardoso, Nuno and Vilela, Rui},
|
31 |
+
booktitle={quot; In Nicoletta Calzolari; Khalid Choukri; Aldo Gangemi; Bente Maegaard; Joseph Mariani; Jan Odjik; Daniel Tapias (ed) Proceedings of the 5 th International Conference on Language Resources and Evaluation (LREC'2006)(Genoa Italy 22-28 May 2006)},
|
32 |
+
year={2006}
|
33 |
+
}
|
34 |
+
"""
|
35 |
+
|
36 |
+
_DESCRIPTION = """
|
37 |
+
The HAREM is a Portuguese language corpus commonly used for Named Entity Recognition tasks. It includes about 93k words, from 129 different texts,
|
38 |
+
from several genres, and language varieties. The split of this dataset version follows the division made by [1], where 7% HAREM
|
39 |
+
documents are the validation set and the miniHAREM corpus (with about 65k words) is the test set. There are two versions of the dataset set,
|
40 |
+
a version that has a total of 10 different named entity classes (Person, Organization, Location, Value, Date, Title, Thing, Event,
|
41 |
+
Abstraction, and Other) and a "selective" version with only 5 classes (Person, Organization, Location, Value, and Date).
|
42 |
+
|
43 |
+
It's important to note that the original version of the HAREM dataset has 2 levels of NER details, namely "Category" and "Sub-type".
|
44 |
+
The dataset version processed here ONLY USE the "Category" level of the original dataset.
|
45 |
+
|
46 |
+
[1] Souza, Fábio, Rodrigo Nogueira, and Roberto Lotufo. "BERTimbau: Pretrained BERT Models for Brazilian Portuguese." Brazilian Conference on Intelligent Systems. Springer, Cham, 2020.
|
47 |
+
"""
|
48 |
+
|
49 |
+
_HOMEPAGE = "https://www.linguateca.pt/primeiroHAREM/harem_coleccaodourada_en.html"
|
50 |
+
|
51 |
+
_LICENSE = ""
|
52 |
+
|
53 |
+
_URLs = {
|
54 |
+
"default": {
|
55 |
+
"train": "https://raw.githubusercontent.com/neuralmind-ai/portuguese-bert/master/ner_evaluation/data/FirstHAREM-total-train.json",
|
56 |
+
"dev": "https://raw.githubusercontent.com/neuralmind-ai/portuguese-bert/master/ner_evaluation/data/FirstHAREM-total-dev.json",
|
57 |
+
"test": "https://raw.githubusercontent.com/neuralmind-ai/portuguese-bert/master/ner_evaluation/data/MiniHAREM-total.json",
|
58 |
+
},
|
59 |
+
"selective": {
|
60 |
+
"train": "https://raw.githubusercontent.com/neuralmind-ai/portuguese-bert/master/ner_evaluation/data/FirstHAREM-selective-train.json",
|
61 |
+
"dev": "https://raw.githubusercontent.com/neuralmind-ai/portuguese-bert/master/ner_evaluation/data/FirstHAREM-selective-dev.json",
|
62 |
+
"test": "https://raw.githubusercontent.com/neuralmind-ai/portuguese-bert/master/ner_evaluation/data/MiniHAREM-selective.json",
|
63 |
+
},
|
64 |
+
}
|
65 |
+
|
66 |
+
|
67 |
+
# method extracted from https://github.com/huggingface/transformers/blob/master/src/transformers/tokenization_utils.py#L77-L89
|
68 |
+
def _is_punctuation(char):
|
69 |
+
"""Checks whether `char` is a punctuation character."""
|
70 |
+
cp = ord(char)
|
71 |
+
# We treat all non-letter/number ASCII as punctuation.
|
72 |
+
# Characters such as "^", "$", and "`" are not in the Unicode
|
73 |
+
# Punctuation class but we treat them as punctuation anyways, for
|
74 |
+
# consistency.
|
75 |
+
if (cp >= 33 and cp <= 47) or (cp >= 58 and cp <= 64) or (cp >= 91 and cp <= 96) or (cp >= 123 and cp <= 126):
|
76 |
+
return True
|
77 |
+
cat = unicodedata.category(char)
|
78 |
+
if cat.startswith("P"):
|
79 |
+
return True
|
80 |
+
return False
|
81 |
+
|
82 |
+
|
83 |
+
# method extracted from https://github.com/huggingface/transformers/blob/master/src/transformers/tokenization_utils.py#L53-L62
|
84 |
+
def _is_whitespace(char):
|
85 |
+
"""Checks whether `char` is a whitespace character."""
|
86 |
+
# \t, \n, and \r are technically control characters but we treat them
|
87 |
+
# as whitespace since they are generally considered as such.
|
88 |
+
if char == " " or char == "\t" or char == "\n" or char == "\r":
|
89 |
+
return True
|
90 |
+
cat = unicodedata.category(char)
|
91 |
+
if cat == "Zs":
|
92 |
+
return True
|
93 |
+
return False
|
94 |
+
|
95 |
+
|
96 |
+
class Token:
|
97 |
+
"""Info about a single token."""
|
98 |
+
|
99 |
+
def __init__(self, text: str, tail: str = ""):
|
100 |
+
|
101 |
+
if not isinstance(text, str) or not text:
|
102 |
+
raise TypeError("text should be a non-empty string.")
|
103 |
+
self.text = text
|
104 |
+
self.tail = tail
|
105 |
+
|
106 |
+
def __len__(self):
|
107 |
+
return len(self.text) + len(self.tail)
|
108 |
+
|
109 |
+
def __add__(self, char):
|
110 |
+
self.text += char
|
111 |
+
return self
|
112 |
+
|
113 |
+
|
114 |
+
def reconstruct_text_from_tokens(tokens: List[Token], include_last_tail: bool = False) -> str:
|
115 |
+
"""Concatenates the text of a sequence of tokens."""
|
116 |
+
|
117 |
+
def text_generator(tokens):
|
118 |
+
for i, token in enumerate(tokens):
|
119 |
+
yield token.text
|
120 |
+
if i < len(tokens) - 1 or include_last_tail:
|
121 |
+
yield token.tail
|
122 |
+
|
123 |
+
return "".join(piece for piece in text_generator(tokens))
|
124 |
+
|
125 |
+
|
126 |
+
def tokenize(text: str) -> Tuple[List[Token], List[int]]:
|
127 |
+
""" Perform whitespace and punctuation tokenization keeping track of char alignment"""
|
128 |
+
doc_tokens = []
|
129 |
+
char_to_word_offset = []
|
130 |
+
|
131 |
+
new_word = True
|
132 |
+
curr_token = None
|
133 |
+
|
134 |
+
def begin_new_token(doc_tokens, text):
|
135 |
+
token = Token(text=text)
|
136 |
+
doc_tokens.append(token)
|
137 |
+
return token
|
138 |
+
|
139 |
+
for offset, c in enumerate(text):
|
140 |
+
if _is_whitespace(c):
|
141 |
+
new_word = True
|
142 |
+
if curr_token:
|
143 |
+
curr_token.tail += c
|
144 |
+
else:
|
145 |
+
if _is_punctuation(c):
|
146 |
+
curr_token = begin_new_token(doc_tokens, c)
|
147 |
+
new_word = True
|
148 |
+
else:
|
149 |
+
if new_word:
|
150 |
+
curr_token = begin_new_token(doc_tokens, c)
|
151 |
+
else:
|
152 |
+
curr_token += c
|
153 |
+
new_word = False
|
154 |
+
|
155 |
+
# OBS: Whitespaces that appear before any tokens will have offset -1
|
156 |
+
# char_to_word_offset.append(len(doc_tokens) - 1)
|
157 |
+
char_to_word_offset.append(max(0, len(doc_tokens) - 1))
|
158 |
+
|
159 |
+
return doc_tokens, char_to_word_offset
|
160 |
+
|
161 |
+
|
162 |
+
class HAREM(datasets.GeneratorBasedBuilder):
|
163 |
+
"""HAREM dataset."""
|
164 |
+
|
165 |
+
VERSION = datasets.Version("1.0.0")
|
166 |
+
|
167 |
+
BUILDER_CONFIGS = [
|
168 |
+
datasets.BuilderConfig(
|
169 |
+
name="default",
|
170 |
+
version=VERSION,
|
171 |
+
description="All the tags (PESSOA, ORGANIZACAO, LOCAL, TEMPO, VALOR, ABSTRACCAO, ACONTECIMENTO, COISA, OBRA, OUTRO) will be used",
|
172 |
+
),
|
173 |
+
datasets.BuilderConfig(
|
174 |
+
name="selective",
|
175 |
+
version=VERSION,
|
176 |
+
description="Only a subset of the tags (PESSOA, ORGANIZACAO, LOCAL, TEMPO, VALOR) will be used",
|
177 |
+
),
|
178 |
+
]
|
179 |
+
|
180 |
+
DEFAULT_CONFIG_NAME = "default"
|
181 |
+
|
182 |
+
def _info(self):
|
183 |
+
|
184 |
+
tags = [
|
185 |
+
"O",
|
186 |
+
"B-PESSOA",
|
187 |
+
"I-PESSOA",
|
188 |
+
"B-ORGANIZACAO",
|
189 |
+
"I-ORGANIZACAO",
|
190 |
+
"B-LOCAL",
|
191 |
+
"I-LOCAL",
|
192 |
+
"B-TEMPO",
|
193 |
+
"I-TEMPO",
|
194 |
+
"B-VALOR",
|
195 |
+
"I-VALOR",
|
196 |
+
]
|
197 |
+
|
198 |
+
if self.config.name == "default":
|
199 |
+
tags += [
|
200 |
+
"B-ABSTRACCAO",
|
201 |
+
"I-ABSTRACCAO",
|
202 |
+
"B-ACONTECIMENTO",
|
203 |
+
"I-ACONTECIMENTO",
|
204 |
+
"B-COISA",
|
205 |
+
"I-COISA",
|
206 |
+
"B-OBRA",
|
207 |
+
"I-OBRA",
|
208 |
+
"B-OUTRO",
|
209 |
+
"I-OUTRO",
|
210 |
+
]
|
211 |
+
|
212 |
+
features = datasets.Features(
|
213 |
+
{
|
214 |
+
"id": datasets.Value("string"),
|
215 |
+
"tokens": datasets.Sequence(datasets.Value("string")),
|
216 |
+
"ner_tags": datasets.Sequence(datasets.features.ClassLabel(names=tags)),
|
217 |
+
}
|
218 |
+
)
|
219 |
+
|
220 |
+
return datasets.DatasetInfo(
|
221 |
+
description=_DESCRIPTION,
|
222 |
+
features=features,
|
223 |
+
supervised_keys=None,
|
224 |
+
homepage=_HOMEPAGE,
|
225 |
+
citation=_CITATION,
|
226 |
+
)
|
227 |
+
|
228 |
+
def _split_generators(self, dl_manager):
|
229 |
+
"""Returns SplitGenerators."""
|
230 |
+
|
231 |
+
my_urls = _URLs[self.config.name]
|
232 |
+
data_dir = dl_manager.download_and_extract(my_urls)
|
233 |
+
|
234 |
+
return [
|
235 |
+
datasets.SplitGenerator(
|
236 |
+
name=datasets.Split.TRAIN,
|
237 |
+
gen_kwargs={"filepath": data_dir["train"], "split": "train"},
|
238 |
+
),
|
239 |
+
datasets.SplitGenerator(
|
240 |
+
name=datasets.Split.TEST,
|
241 |
+
gen_kwargs={"filepath": data_dir["test"], "split": "test"},
|
242 |
+
),
|
243 |
+
datasets.SplitGenerator(
|
244 |
+
name=datasets.Split.VALIDATION,
|
245 |
+
gen_kwargs={"filepath": data_dir["dev"], "split": "dev"},
|
246 |
+
),
|
247 |
+
]
|
248 |
+
|
249 |
+
def _generate_examples(self, filepath, split):
|
250 |
+
""" Yields examples. """
|
251 |
+
|
252 |
+
logging.info("⏳ Generating examples from = %s", filepath)
|
253 |
+
|
254 |
+
with open(filepath, "r", encoding="utf-8") as f:
|
255 |
+
|
256 |
+
input_data = json.load(f)
|
257 |
+
id_ = 0
|
258 |
+
|
259 |
+
for document in input_data:
|
260 |
+
doc_text = document["doc_text"]
|
261 |
+
doc_id = document["doc_id"]
|
262 |
+
|
263 |
+
doc_tokens, char_to_word_offset = tokenize(doc_text)
|
264 |
+
tags = ["O"] * len(doc_tokens)
|
265 |
+
|
266 |
+
def set_label(index, tag):
|
267 |
+
if tags[index] != "O":
|
268 |
+
logging.warning(
|
269 |
+
"Overwriting tag %s at position %s to %s",
|
270 |
+
tags[index],
|
271 |
+
index,
|
272 |
+
tag,
|
273 |
+
)
|
274 |
+
tags[index] = tag
|
275 |
+
|
276 |
+
for entity in document["entities"]:
|
277 |
+
entity_text = entity["text"]
|
278 |
+
entity_type = entity["label"]
|
279 |
+
start_token = None
|
280 |
+
end_token = None
|
281 |
+
|
282 |
+
entity_start_offset = entity["start_offset"]
|
283 |
+
entity_end_offset = entity["end_offset"]
|
284 |
+
start_token = char_to_word_offset[entity_start_offset]
|
285 |
+
|
286 |
+
# end_offset is NOT inclusive to the text, e.g.,
|
287 |
+
# entity_text == doc_text[start_offset:end_offset]
|
288 |
+
end_token = char_to_word_offset[entity_end_offset - 1]
|
289 |
+
|
290 |
+
assert start_token <= end_token, "End token cannot come before start token."
|
291 |
+
reconstructed_text = reconstruct_text_from_tokens(doc_tokens[start_token : (end_token + 1)])
|
292 |
+
assert (
|
293 |
+
entity_text.strip() == reconstructed_text
|
294 |
+
), "Entity text and reconstructed text are not equal: %s != %s" % (
|
295 |
+
entity_text,
|
296 |
+
reconstructed_text,
|
297 |
+
)
|
298 |
+
|
299 |
+
for token_index in range(start_token, end_token + 1):
|
300 |
+
if token_index == start_token:
|
301 |
+
tag = "B-" + entity_type
|
302 |
+
else:
|
303 |
+
tag = "I-" + entity_type
|
304 |
+
set_label(token_index, tag)
|
305 |
+
|
306 |
+
yield id_, {
|
307 |
+
"id": doc_id,
|
308 |
+
"tokens": [x.text for x in doc_tokens],
|
309 |
+
"ner_tags": tags,
|
310 |
+
}
|
311 |
+
id_ += 1
|