Hugging Face's logo

Datasets:
relbert
/
nell

Modalities:
Text
Languages:
English
Size:
1K - 10K
Libraries:
Datasets
License:
Dataset card Data Studio Files Community
asahi417 commited on
Commit
bb828f8
·
1 Parent(s): a5cba06

init

Browse files
Files changed (3) hide show
  1. README.md +111 -51
  2. fewshot_link_prediction.py +2 -2
  3. stats.py +5 -1
README.md CHANGED
@@ -35,75 +35,76 @@ An example of `test` of `nell` looks as follows.
35
 
36
  ## Statistics on the NELL test split
37
 
38
- - Entity Types
 
39
 
40
  | entity_type | tail | head |
41
  |:-------------------------|-------:|-------:|
42
- | geopoliticallocation | 14 | 24 |
43
- | crustacean | 25 | 11 |
44
- | amphibian | 0 | 2 |
45
- | criminal | 1 | 0 |
46
- | arthropod | 41 | 32 |
47
  | videogame | 0 | 4 |
48
- | mlsoftware | 0 | 1 |
49
- | animal | 30 | 97 |
50
- | politician | 58 | 23 |
51
  | organization | 2 | 32 |
52
- | insect | 270 | 230 |
53
- | coach | 245 | 1 |
54
- | personmexico | 20 | 13 |
55
- | personus | 6 | 1 |
56
- | automobilemaker | 54 | 274 |
57
- | chemical | 0 | 1 |
58
  | astronaut | 1 | 0 |
 
 
 
 
 
 
59
  | company | 147 | 1 |
60
- | arachnid | 6 | 1 |
61
- | software | 0 | 42 |
62
- | biotechcompany | 11 | 0 |
63
- | mammal | 0 | 23 |
64
  | product | 0 | 62 |
65
- | sportsgame | 0 | 74 |
 
 
 
66
  | sport | 74 | 93 |
67
- | agriculturalproduct | 0 | 87 |
68
- | location | 2 | 0 |
69
- | hobby | 0 | 10 |
70
- | vehicle | 0 | 2 |
71
- | director | 1 | 0 |
72
  | planet | 1 | 0 |
 
 
 
73
  | athlete | 59 | 34 |
74
- | food | 0 | 4 |
75
- | grain | 0 | 2 |
76
- | politicianus | 360 | 352 |
77
  | automobilemodel | 0 | 100 |
78
  | coffeedrink | 0 | 11 |
79
- | city | 161 | 42 |
80
- | person | 96 | 14 |
81
- | bodypart | 69 | 0 |
82
- | female | 3 | 3 |
83
- | journalist | 1 | 0 |
84
- | geopoliticalorganization | 17 | 1 |
85
- | visualizablescene | 3 | 3 |
86
- | professor | 0 | 1 |
87
- | reptile | 0 | 4 |
88
- | drug | 0 | 1 |
89
- | stateorprovince | 0 | 38 |
90
- | beverage | 0 | 27 |
91
- | school | 0 | 11 |
92
- | county | 4 | 10 |
93
- | male | 5 | 3 |
94
  | celebrity | 2 | 4 |
95
- | country | 317 | 27 |
 
 
 
 
 
 
 
96
  | personaustralia | 5 | 4 |
 
 
 
 
 
 
97
  | sportsteam | 0 | 295 |
98
- | fruit | 0 | 1 |
 
 
 
 
 
 
 
 
 
 
99
  | island | 0 | 1 |
100
- | invertebrate | 43 | 14 |
101
- | personnorthamerica | 3 | 1 |
102
- | personeurope | 1 | 0 |
103
- | vegetable | 0 | 8 |
104
- | legume | 0 | 1 |
105
 
106
- - Relation Types
107
 
108
  | relation | relation |
109
  |:-----------------------------------------------|-----------:|
@@ -119,6 +120,65 @@ An example of `test` of `nell` looks as follows.
119
  | concept:sportsgamesport | 74 |
120
  | concept:teamcoach | 341 |
121
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
122
 
123
  ### Citation Information
124
  ```
 
35
 
36
  ## Statistics on the NELL test split
37
 
38
+
39
+ - Entity Types (nell)
40
 
41
  | entity_type | tail | head |
42
  |:-------------------------|-------:|-------:|
 
 
 
 
 
43
  | videogame | 0 | 4 |
44
+ | crustacean | 25 | 11 |
 
 
45
  | organization | 2 | 32 |
46
+ | invertebrate | 43 | 14 |
47
+ | agriculturalproduct | 0 | 87 |
 
 
 
 
48
  | astronaut | 1 | 0 |
49
+ | vegetable | 0 | 8 |
50
+ | geopoliticallocation | 14 | 24 |
51
+ | person | 96 | 14 |
52
+ | drug | 0 | 1 |
53
+ | arthropod | 41 | 32 |
54
+ | location | 2 | 0 |
55
  | company | 147 | 1 |
56
+ | female | 3 | 3 |
 
 
 
57
  | product | 0 | 62 |
58
+ | chemical | 0 | 1 |
59
+ | legume | 0 | 1 |
60
+ | county | 4 | 10 |
61
+ | mlsoftware | 0 | 1 |
62
  | sport | 74 | 93 |
63
+ | fruit | 0 | 1 |
64
+ | personeurope | 1 | 0 |
 
 
 
65
  | planet | 1 | 0 |
66
+ | mammal | 0 | 23 |
67
+ | professor | 0 | 1 |
68
+ | criminal | 1 | 0 |
69
  | athlete | 59 | 34 |
70
+ | insect | 270 | 230 |
71
+ | school | 0 | 11 |
 
72
  | automobilemodel | 0 | 100 |
73
  | coffeedrink | 0 | 11 |
74
+ | food | 0 | 4 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
75
  | celebrity | 2 | 4 |
76
+ | biotechcompany | 11 | 0 |
77
+ | animal | 30 | 97 |
78
+ | visualizablescene | 3 | 3 |
79
+ | politician | 58 | 23 |
80
+ | software | 0 | 42 |
81
+ | sportsgame | 0 | 74 |
82
+ | grain | 0 | 2 |
83
+ | city | 161 | 42 |
84
  | personaustralia | 5 | 4 |
85
+ | politicianus | 360 | 352 |
86
+ | hobby | 0 | 10 |
87
+ | arachnid | 6 | 1 |
88
+ | country | 317 | 27 |
89
+ | male | 5 | 3 |
90
+ | personnorthamerica | 3 | 1 |
91
  | sportsteam | 0 | 295 |
92
+ | stateorprovince | 0 | 38 |
93
+ | personus | 6 | 1 |
94
+ | coach | 245 | 1 |
95
+ | director | 1 | 0 |
96
+ | automobilemaker | 54 | 274 |
97
+ | reptile | 0 | 4 |
98
+ | journalist | 1 | 0 |
99
+ | vehicle | 0 | 2 |
100
+ | bodypart | 69 | 0 |
101
+ | amphibian | 0 | 2 |
102
+ | beverage | 0 | 27 |
103
  | island | 0 | 1 |
104
+ | geopoliticalorganization | 17 | 1 |
105
+ | personmexico | 20 | 13 |
 
 
 
106
 
107
+ - Relation Types (nell)
108
 
109
  | relation | relation |
110
  |:-----------------------------------------------|-----------:|
 
120
  | concept:sportsgamesport | 74 |
121
  | concept:teamcoach | 341 |
122
 
123
+ - Vocab Size (nell): 68544
124
+
125
+ - Entity Types (nell_filter)
126
+
127
+ | entity_type | tail | head |
128
+ |:-------------------------|-------:|-------:|
129
+ | videogame | 0 | 4 |
130
+ | organization | 2 | 32 |
131
+ | astronaut | 1 | 0 |
132
+ | geopoliticallocation | 8 | 24 |
133
+ | person | 96 | 0 |
134
+ | drug | 0 | 1 |
135
+ | company | 144 | 1 |
136
+ | female | 3 | 3 |
137
+ | product | 0 | 62 |
138
+ | county | 4 | 10 |
139
+ | personeurope | 1 | 0 |
140
+ | planet | 1 | 0 |
141
+ | professor | 0 | 1 |
142
+ | criminal | 1 | 0 |
143
+ | athlete | 59 | 0 |
144
+ | school | 0 | 11 |
145
+ | automobilemodel | 0 | 100 |
146
+ | celebrity | 2 | 4 |
147
+ | biotechcompany | 10 | 0 |
148
+ | visualizablescene | 3 | 3 |
149
+ | politician | 58 | 23 |
150
+ | software | 0 | 42 |
151
+ | city | 161 | 42 |
152
+ | personaustralia | 5 | 0 |
153
+ | politicianus | 360 | 352 |
154
+ | country | 91 | 27 |
155
+ | male | 5 | 1 |
156
+ | personnorthamerica | 3 | 0 |
157
+ | sportsteam | 0 | 295 |
158
+ | stateorprovince | 0 | 38 |
159
+ | personus | 6 | 1 |
160
+ | coach | 245 | 0 |
161
+ | director | 1 | 0 |
162
+ | automobilemaker | 54 | 273 |
163
+ | journalist | 1 | 0 |
164
+ | island | 0 | 1 |
165
+ | geopoliticalorganization | 7 | 1 |
166
+ | personmexico | 20 | 0 |
167
+
168
+ - Relation Types (nell_filter)
169
+
170
+ | relation | relation |
171
+ |:-----------------------------------------------|-----------:|
172
+ | concept:automobilemakerdealersincity | 177 |
173
+ | concept:automobilemakerdealersincountry | 96 |
174
+ | concept:geopoliticallocationresidenceofpersion | 143 |
175
+ | concept:politicianusendorsespoliticianus | 386 |
176
+ | concept:producedby | 209 |
177
+ | concept:teamcoach | 341 |
178
+
179
+ Vocab Size (nell_filter)
180
+ 53887
181
+
182
 
183
  ### Citation Information
184
  ```
fewshot_link_prediction.py CHANGED
@@ -5,7 +5,7 @@ import datasets
5
  logger = datasets.logging.get_logger(__name__)
6
  _DESCRIPTION = """Few shots link prediction dataset. """
7
  _NAME = "fewshot_link_prediction"
8
- _VERSION = "0.0.3"
9
  _CITATION = """
10
  @inproceedings{xiong-etal-2018-one,
11
  title = "One-Shot Relational Learning for Knowledge Graphs",
@@ -28,7 +28,7 @@ _CITATION = """
28
 
29
  _HOME_PAGE = "https://github.com/asahi417/relbert"
30
  _URL = f'https://huggingface.co/datasets/relbert/{_NAME}/resolve/main/data'
31
- _TYPES = ["nell", "nell_filter", "wiki"]
32
  _URLS = {i: {
33
  str(datasets.Split.TRAIN): [f'{_URL}/{i}.train.jsonl'],
34
  str(datasets.Split.VALIDATION): [f'{_URL}/{i}.validation.jsonl'],
 
5
  logger = datasets.logging.get_logger(__name__)
6
  _DESCRIPTION = """Few shots link prediction dataset. """
7
  _NAME = "fewshot_link_prediction"
8
+ _VERSION = "0.0.4"
9
  _CITATION = """
10
  @inproceedings{xiong-etal-2018-one,
11
  title = "One-Shot Relational Learning for Knowledge Graphs",
 
28
 
29
  _HOME_PAGE = "https://github.com/asahi417/relbert"
30
  _URL = f'https://huggingface.co/datasets/relbert/{_NAME}/resolve/main/data'
31
+ _TYPES = ["nell_filter", "nell", "wiki"]
32
  _URLS = {i: {
33
  str(datasets.Split.TRAIN): [f'{_URL}/{i}.train.jsonl'],
34
  str(datasets.Split.VALIDATION): [f'{_URL}/{i}.validation.jsonl'],
stats.py CHANGED
@@ -2,7 +2,7 @@ import pandas as pd
2
  from datasets import load_dataset
3
 
4
  for _type in ['nell', 'nell_filter']:
5
- data = load_dataset("fewshot_link_prediction", _type, split='test')
6
  df = data.to_pandas()
7
 
8
  print(f"\nEntity Types ({_type})")
@@ -15,4 +15,8 @@ for _type in ['nell', 'nell_filter']:
15
  print(f"\nRelation Types ({_type})")
16
  print(df.groupby("relation")['relation'].count().to_markdown())
17
 
 
 
 
 
18
 
 
2
  from datasets import load_dataset
3
 
4
  for _type in ['nell', 'nell_filter']:
5
+ data = load_dataset("relbert/fewshot_link_prediction", _type, split='test')
6
  df = data.to_pandas()
7
 
8
  print(f"\nEntity Types ({_type})")
 
15
  print(f"\nRelation Types ({_type})")
16
  print(df.groupby("relation")['relation'].count().to_markdown())
17
 
18
+ print(f"\nVocab Size ({_type})")
19
+ with open(f"data/{_type}.vocab.txt") as f:
20
+ length = len([i for i in f.read().split('\n') if len(i) > 0])
21
+ print(length)
22