asahi417 commited on
Commit
0bf19bb
1 Parent(s): 5cdadef
Files changed (1) hide show
  1. README.md +28 -61
README.md CHANGED
@@ -11,21 +11,23 @@ task_categories:
11
  - token-classification
12
  task_ids:
13
  - named-entity-recognition
14
- pretty_name: Ontonotes5
15
  ---
16
 
17
- # Dataset Card for "tner/ontonotes5"
18
 
19
  ## Dataset Description
20
 
21
  - **Repository:** [T-NER](https://github.com/asahi417/tner)
22
- - **Paper:** [https://aclanthology.org/N06-2015/](https://aclanthology.org/N06-2015/)
23
- - **Dataset:** Ontonotes5
24
- - **Domain:** News
25
- - **Number of Entity:** `8 (`CARDINAL`, `DATE`, `PERSON`, `NORP`, `GPE`, `LAW`, `PERCENT`, `ORDINAL`, `MONEY`, `WORK_OF_ART`, `FAC`, `TIME`, `QUANTITY`, `PRODUCT`, `LANGUAGE`, `ORG`, `LOC`, `EVENT`)
26
 
27
  ### Dataset Summary
28
- Ontonotes5 NER dataset formatted in a part of [TNER](https://github.com/asahi417/tner) project.
 
 
29
 
30
  ## Dataset Structure
31
 
@@ -34,52 +36,20 @@ An example of `train` looks as follows.
34
 
35
  ```
36
  {
37
- 'tags': [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 4, 5, 0, 0, 0, 0, 11, 12, 12, 12, 12, 0, 0, 7, 0, 0, 0, 0, 0],
38
- 'tokens': ['``', 'It', "'s", 'very', 'costly', 'and', 'time', '-', 'consuming', ',', "''", 'says', 'Phil', 'Rosen', ',', 'a', 'partner', 'in', 'Fleet', '&', 'Leasing', 'Management', 'Inc.', ',', 'a', 'Boston', 'car', '-', 'leasing', 'company', '.']
39
  }
40
  ```
41
 
42
  ### Label ID
43
- The label2id dictionary can be found at [here](https://huggingface.co/datasets/tner/onotonotes5/raw/main/dataset/label.json).
44
  ```python
45
  {
46
  "O": 0,
47
- "B-CARDINAL": 1,
48
- "B-DATE": 2,
49
- "I-DATE": 3,
50
- "B-PERSON": 4,
51
- "I-PERSON": 5,
52
- "B-NORP": 6,
53
- "B-GPE": 7,
54
- "I-GPE": 8,
55
- "B-LAW": 9,
56
- "I-LAW": 10,
57
- "B-ORG": 11,
58
- "I-ORG": 12,
59
- "B-PERCENT": 13,
60
- "I-PERCENT": 14,
61
- "B-ORDINAL": 15,
62
- "B-MONEY": 16,
63
- "I-MONEY": 17,
64
- "B-WORK_OF_ART": 18,
65
- "I-WORK_OF_ART": 19,
66
- "B-FAC": 20,
67
- "B-TIME": 21,
68
- "I-CARDINAL": 22,
69
- "B-LOC": 23,
70
- "B-QUANTITY": 24,
71
- "I-QUANTITY": 25,
72
- "I-NORP": 26,
73
- "I-LOC": 27,
74
- "B-PRODUCT": 28,
75
- "I-TIME": 29,
76
- "B-EVENT": 30,
77
- "I-EVENT": 31,
78
- "I-FAC": 32,
79
- "B-LANGUAGE": 33,
80
- "I-PRODUCT": 34,
81
- "I-ORDINAL": 35,
82
- "I-LANGUAGE": 36
83
  }
84
  ```
85
 
@@ -87,24 +57,21 @@ The label2id dictionary can be found at [here](https://huggingface.co/datasets/t
87
 
88
  | name |train|validation|test|
89
  |---------|----:|---------:|---:|
90
- |ontonotes5|59924| 8528|8262|
91
 
92
  ### Citation Information
93
 
94
  ```
95
- @inproceedings{hovy-etal-2006-ontonotes,
96
- title = "{O}nto{N}otes: The 90{\%} Solution",
97
- author = "Hovy, Eduard and
98
- Marcus, Mitchell and
99
- Palmer, Martha and
100
- Ramshaw, Lance and
101
- Weischedel, Ralph",
102
- booktitle = "Proceedings of the Human Language Technology Conference of the {NAACL}, Companion Volume: Short Papers",
103
- month = jun,
104
- year = "2006",
105
- address = "New York City, USA",
106
- publisher = "Association for Computational Linguistics",
107
- url = "https://aclanthology.org/N06-2015",
108
- pages = "57--60",
109
  }
110
  ```
 
11
  - token-classification
12
  task_ids:
13
  - named-entity-recognition
14
+ pretty_name: FIN
15
  ---
16
 
17
+ # Dataset Card for "tner/fin"
18
 
19
  ## Dataset Description
20
 
21
  - **Repository:** [T-NER](https://github.com/asahi417/tner)
22
+ - **Paper:** [https://aclanthology.org/U15-1010.pdf](https://aclanthology.org/U15-1010.pdf)
23
+ - **Dataset:** FIN
24
+ - **Domain:** Financial News
25
+ - **Number of Entity:** 4 (`ORG`, `LOC`, `PER`, `MISC`)
26
 
27
  ### Dataset Summary
28
+ FIN NER dataset formatted in a part of [TNER](https://github.com/asahi417/tner) project.
29
+ Original FIN dataset contains two variants of datasets, FIN3 and FIN5 where the FIN3 is the test set, while FIN5 is the training set.
30
+ We take same amount of instances randomly from the training set and create a validation set with the subset.
31
 
32
  ## Dataset Structure
33
 
 
36
 
37
  ```
38
  {
39
+ "tags": [0, 0, 0, 0, 0, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
40
+ "tokens": ["1", ".", "1", ".", "4", "Borrower", "engages", "in", "criminal", "conduct", "or", "is", "involved", "in", "criminal", "activities", ";"]
41
  }
42
  ```
43
 
44
  ### Label ID
45
+ The label2id dictionary can be found at [here](https://huggingface.co/datasets/tner/fin/raw/main/dataset/label.json).
46
  ```python
47
  {
48
  "O": 0,
49
+ "I-ORG": 1,
50
+ "I-LOC": 2,
51
+ "I-PER": 3,
52
+ "I-MISC": 4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
53
  }
54
  ```
55
 
 
57
 
58
  | name |train|validation|test|
59
  |---------|----:|---------:|---:|
60
+ |fin |861 | 303| 303|
61
 
62
  ### Citation Information
63
 
64
  ```
65
+ @inproceedings{salinas-alvarado-etal-2015-domain,
66
+ title = "Domain Adaption of Named Entity Recognition to Support Credit Risk Assessment",
67
+ author = "Salinas Alvarado, Julio Cesar and
68
+ Verspoor, Karin and
69
+ Baldwin, Timothy",
70
+ booktitle = "Proceedings of the Australasian Language Technology Association Workshop 2015",
71
+ month = dec,
72
+ year = "2015",
73
+ address = "Parramatta, Australia",
74
+ url = "https://aclanthology.org/U15-1010",
75
+ pages = "84--90",
 
 
 
76
  }
77
  ```