fix readme
Browse files- .gitattributes +1 -0
- .gitignore +1 -0
- README.md +8 -7
- check_predicate.py +0 -11
- data/t_rex.filter.jsonl +2 -2
- data/t_rex.filter_unified.jsonl +3 -0
- filtering_denoise.py +0 -47
- predicate_manual_check.csv +660 -0
- process.py +78 -24
- t_rex.py +5 -4
- unify_predicate.py +69 -0
.gitattributes
CHANGED
@@ -79,3 +79,4 @@ data/t_rex.filter.min_entity_12_max_predicate_10.jsonl filter=lfs diff=lfs merge
|
|
79 |
data/t_rex.filter.min_entity_16_max_predicate_100.jsonl filter=lfs diff=lfs merge=lfs -text
|
80 |
data/t_rex.filter.min_entity_4_max_predicate_25.jsonl filter=lfs diff=lfs merge=lfs -text
|
81 |
data/t_rex.filter.min_entity_8_max_predicate_25.jsonl filter=lfs diff=lfs merge=lfs -text
|
|
|
|
79 |
data/t_rex.filter.min_entity_16_max_predicate_100.jsonl filter=lfs diff=lfs merge=lfs -text
|
80 |
data/t_rex.filter.min_entity_4_max_predicate_25.jsonl filter=lfs diff=lfs merge=lfs -text
|
81 |
data/t_rex.filter.min_entity_8_max_predicate_25.jsonl filter=lfs diff=lfs merge=lfs -text
|
82 |
+
data/t_rex.filter_unified.jsonl filter=lfs diff=lfs merge=lfs -text
|
.gitignore
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
data_raw
|
README.md
CHANGED
@@ -21,12 +21,13 @@ We split the raw T-REX dataset into train/validation/test split by the ratio of
|
|
21 |
### Filtering to Remove Noise
|
22 |
|
23 |
We apply filtering to keep triples with alpha-numeric subject and object, as well as triples with at least either of subject or object is a named-entity.
|
|
|
24 |
|
25 |
-
| Dataset
|
26 |
-
|
27 |
-
| Triples
|
28 |
-
| Predicate| 931 |
|
29 |
-
| Entity
|
30 |
|
31 |
### Filtering to Purify the Dataset
|
32 |
We reduce the size of the dataset by applying filtering based on the number of predicates and entities in the triples.
|
@@ -54,11 +55,11 @@ we choose top-`max predicate` triples based on the frequency of the subject and
|
|
54 |
|
55 |
- distribution of entities
|
56 |
|
57 |
-
<img src="https://huggingface.co/datasets/relbert/t_rex/resolve/main/data/stats.entity_distribution.png" alt="" width="
|
58 |
|
59 |
- distribution of predicates
|
60 |
|
61 |
-
<img src="https://huggingface.co/datasets/relbert/t_rex/resolve/main/data/stats.predicate_distribution.png" alt="" width="
|
62 |
|
63 |
|
64 |
## Dataset Structure
|
|
|
21 |
### Filtering to Remove Noise
|
22 |
|
23 |
We apply filtering to keep triples with alpha-numeric subject and object, as well as triples with at least either of subject or object is a named-entity.
|
24 |
+
After the filtering, we manually remove too vague and noisy predicate, and unify same predicates with different names.
|
25 |
|
26 |
+
| Dataset | Raw | Filter | Unification |
|
27 |
+
|----------:|----------:|----------:|--------------:|
|
28 |
+
| Triples | 941,663 | 583,333 | 432,795 |
|
29 |
+
| Predicate | 931 | 659 | 247 |
|
30 |
+
| Entity | 270,801 | 197,163 | 149,172 |
|
31 |
|
32 |
### Filtering to Purify the Dataset
|
33 |
We reduce the size of the dataset by applying filtering based on the number of predicates and entities in the triples.
|
|
|
55 |
|
56 |
- distribution of entities
|
57 |
|
58 |
+
<img src="https://huggingface.co/datasets/relbert/t_rex/resolve/main/data/stats.entity_distribution.png" alt="" width="600" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
|
59 |
|
60 |
- distribution of predicates
|
61 |
|
62 |
+
<img src="https://huggingface.co/datasets/relbert/t_rex/resolve/main/data/stats.predicate_distribution.png" alt="" width="600" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
|
63 |
|
64 |
|
65 |
## Dataset Structure
|
check_predicate.py
DELETED
@@ -1,11 +0,0 @@
|
|
1 |
-
import json
|
2 |
-
import pandas as pd
|
3 |
-
|
4 |
-
with open("data/t_rex.filter.jsonl") as f:
|
5 |
-
data = pd.DataFrame([json.loads(i) for i in f.read().split('\n') if len(i) > 0])
|
6 |
-
freq = data.groupby("predicate").count()['title']
|
7 |
-
data['freq'] = [freq.loc[i] for i in data['predicate']]
|
8 |
-
tmp = data.groupby("predicate").sample(10, replace=True)
|
9 |
-
tmp = tmp.drop_duplicates()
|
10 |
-
tmp.to_csv("data/t_rex.filter.predicate_check_sample.csv", index=False)
|
11 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
data/t_rex.filter.jsonl
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ce5fdbf59ad4aaa7de18198863aa1d1b9e451386748f0a854b3b7ab1bff7f79f
|
3 |
+
size 502849464
|
data/t_rex.filter_unified.jsonl
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c0aeaa043780559579cc3326e9eb923e09de1842590add1b1d92e4b19454f586
|
3 |
+
size 408520515
|
filtering_denoise.py
DELETED
@@ -1,47 +0,0 @@
|
|
1 |
-
import json
|
2 |
-
import string
|
3 |
-
import re
|
4 |
-
|
5 |
-
stopwords = ["he", "she", "they", "it"]
|
6 |
-
list_alnum = string.ascii_lowercase + '0123456789 '
|
7 |
-
|
8 |
-
|
9 |
-
def filtering(entry):
|
10 |
-
|
11 |
-
def _subfilter(token):
|
12 |
-
if len(re.findall(rf'[^{list_alnum}]+', token)) != 0:
|
13 |
-
return False
|
14 |
-
if token in stopwords:
|
15 |
-
return False
|
16 |
-
if token.startswith("www"):
|
17 |
-
return False
|
18 |
-
if token.startswith("."):
|
19 |
-
return False
|
20 |
-
if token.startswith(","):
|
21 |
-
return False
|
22 |
-
if token.startswith("$"):
|
23 |
-
return False
|
24 |
-
if token.startswith("+"):
|
25 |
-
return False
|
26 |
-
if token.startswith("#"):
|
27 |
-
return False
|
28 |
-
return True
|
29 |
-
|
30 |
-
if not _subfilter(entry["object"].lower()):
|
31 |
-
return False
|
32 |
-
if not _subfilter(entry["subject"].lower()):
|
33 |
-
return False
|
34 |
-
|
35 |
-
if entry['object'].islower() and entry['subject'].islower():
|
36 |
-
return False
|
37 |
-
|
38 |
-
return True
|
39 |
-
|
40 |
-
|
41 |
-
with open(f"data/t_rex.raw.jsonl") as f:
|
42 |
-
data = [json.loads(i) for i in f.read().split('\n') if len(i) > 0]
|
43 |
-
print(f"[before]: {len(data)}")
|
44 |
-
data = [i for i in data if filtering(i)]
|
45 |
-
print(f"[after] : {len(data)}")
|
46 |
-
with open(f"data/t_rex.filter.jsonl", 'w') as f:
|
47 |
-
f.write('\n'.join([json.dumps(i) for i in data]))
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
predicate_manual_check.csv
ADDED
@@ -0,0 +1,660 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
unique predicates,ok,pretty relation name,pretty relation name is reverse,same as,reverse of,remove (noisy),remove (too vague)
|
2 |
+
abolished,,,,ending,,,
|
3 |
+
academic discipline,,,,discipline,,,
|
4 |
+
studies,x,[Academic Subject] studies [Topic],,,,,
|
5 |
+
actor,,,,film starring,,,
|
6 |
+
actress,,,,film starring,,,
|
7 |
+
signatory,x,[Act] is signed by [Person],,,,,
|
8 |
+
adjacent to,,,,next to,,,
|
9 |
+
administered by,,,,maintained by,,,
|
10 |
+
maiden flight,x,[Aircraft] makes the maiden flight on [Date],,,,,
|
11 |
+
advisor,,,,studied at,,,
|
12 |
+
hub,x,[Airline] has a hub in [Location],,,,,
|
13 |
+
airline alliance,x,[Airline] is in [Airline Alliance],,,,,
|
14 |
+
alma mater,,,,studied at,,,
|
15 |
+
fleet,x,[Army] has [Fleet],,,,,
|
16 |
+
alumni of,,,,studied at,,,
|
17 |
+
alumnus of,,,,studied at,,,
|
18 |
+
amalgamation of,,,,,,x,
|
19 |
+
garrison,x,[Army] is based in [Location],,,,,
|
20 |
+
series,x,[Art Work] follows after [Art Work],,,,,
|
21 |
+
painting of,x,[Art Work] is a painting of [Person],,,,,
|
22 |
+
architect,,,,architectural style,,,
|
23 |
+
translation of,x,[Art Work] is a translation of [Art Work],,,,,
|
24 |
+
arena,,,,home field,,,
|
25 |
+
artery,,,,,,x,
|
26 |
+
artist,,,,,,x,
|
27 |
+
artists,,,,,,x,
|
28 |
+
aspect of,,,,,,x,
|
29 |
+
assassin,,,,murdered by,,,
|
30 |
+
assassinated by,,,,murdered by,,,
|
31 |
+
adapted from,x,[Art Work] is adapted from [Art Work],,,,,
|
32 |
+
assembly of,,,,,assembly,,
|
33 |
+
director,x,[Art Work] is directed by [Person],,,,,
|
34 |
+
painter,x,[Art Work] is painted by [Person],,,,,
|
35 |
+
published in,x,[Art Work] is published in [Magazine],,,,,
|
36 |
+
authority,,,,council,,,
|
37 |
+
published,x,[Art Work] is published on [Date],,,,,
|
38 |
+
awarded,,,,award,,,
|
39 |
+
awarded by,,,,,award,,
|
40 |
+
awards,,,,award,,,
|
41 |
+
sculptor,x,[Art Work] is sculpted by [Person],,,,,
|
42 |
+
band,,,,music by,,,
|
43 |
+
based on,,,,series,,,
|
44 |
+
battle,,,,war,,,
|
45 |
+
began,,,,starting,,,
|
46 |
+
beginning,,,,starting,,,
|
47 |
+
belongs to,,,,,divisions,,
|
48 |
+
bestowed by,,,,founder,,,
|
49 |
+
author,x,[Art Work] is written by [Person],,,,,
|
50 |
+
engine,x,[Artifact] has [Engine],,,,,
|
51 |
+
birthday,,,,birth place,,,
|
52 |
+
birthplace,,,,birth date,,,
|
53 |
+
book by,,,,author,,,
|
54 |
+
border,,,,next to,,,
|
55 |
+
bordered by,,,,next to,,,
|
56 |
+
borders,,,,next to,,,
|
57 |
+
born at,,,,birth place,,,
|
58 |
+
born in,,,,birth place,,,
|
59 |
+
born on,,,,birth date,,,
|
60 |
+
boyfriend,,,,,husband,,
|
61 |
+
branch,,,,sector,,,
|
62 |
+
brand,,,,,product,,
|
63 |
+
shape,x,[Artifact] has a shape of [Shape],,,,,
|
64 |
+
coat of arms,x,[Artifact] is a coat of arms of [Group],x,,,,
|
65 |
+
motif,x,[Artifact] is a motif for [Art Work],x,,,,
|
66 |
+
built by,,,,builder,,,
|
67 |
+
built from,,,,constructed from,,,
|
68 |
+
burial place,,,,tomb,,,
|
69 |
+
buried in,,,,tomb,,,
|
70 |
+
patron saint,x,[Artifact] is a patron saint of [Country],x,,,,
|
71 |
+
capital,,,,,capital of,,
|
72 |
+
capital city,,,,,capital of,,
|
73 |
+
result,x,[Artifact] is a result of [Artifact],,,,,
|
74 |
+
type of,x,[Artifact] is a type of [Type],,,,,
|
75 |
+
career,,,,job,,,
|
76 |
+
edition of,x,[Artifact] is a version of [Artifact],,,,,
|
77 |
+
cartridge,,,,,,x,
|
78 |
+
cast member,,,,film starring,,,
|
79 |
+
established,x,[Artifact] is built on [Date],,,,,
|
80 |
+
cause,,,,caused by,,,
|
81 |
+
cause of,,,,,caused by,,
|
82 |
+
discoverer,x,[Artifact] is discovered by [Person],,,,,
|
83 |
+
formation,x,[Artifact] is formation of [Army],,,,,
|
84 |
+
CEO,,,,chief executive,,,
|
85 |
+
formed from,x,[Artifact] is formed from [Artifact],,,,,
|
86 |
+
chairperson,,,,chairman,,,
|
87 |
+
contained within,x,[Artifact] is found in [Artifact],,,,,
|
88 |
+
colour,x,[Artifact] is in [Color],,,,,
|
89 |
+
channel,,,,broadcast on,,,
|
90 |
+
influenced by,x,[Artifact] is influenced by [Artifact],,,,,
|
91 |
+
listing,x,[Artifact] is listed in [List],,,,,
|
92 |
+
chief executive officer,,,,chief executive,,,
|
93 |
+
children,,,,,father,,
|
94 |
+
maintained by,x,[Artifact] is maintained by [Company],,,,,
|
95 |
+
chromosome,,,,,,x,
|
96 |
+
manufacturer,x,[Artifact] is manufactured by [Company],,,,,
|
97 |
+
citizen of,,,,residence,,,
|
98 |
+
citizenship,,,,residence,,,
|
99 |
+
named for,x,[Artifact] is name of [Artifact],,,,,
|
100 |
+
city council,,,,council,,,
|
101 |
+
named after,x,[Artifact] is named after [Person],,,,,
|
102 |
+
operating system,x,[Artifact] is the OS of [Software],,,,,
|
103 |
+
co-driver,,,,,,x,
|
104 |
+
coach,,,,head coach,,,
|
105 |
+
platform,x,[Artifact] is the platform of [Game],x,,,,
|
106 |
+
used by,x,[Artifact] is used by [Person],,,,,
|
107 |
+
namesake,x,[Artifact] is used for its namesake of [Artifact],,,,,
|
108 |
+
color,,,,colour,,,
|
109 |
+
producer,x,[Artist] is produced by [Person],,,,,
|
110 |
+
movement,x,[Artists] leads [Movement],,,,,
|
111 |
+
commissioned by,,,,producer,,,
|
112 |
+
compiled by,,,,,,x,
|
113 |
+
composed by,,,,music by,,,
|
114 |
+
athletes,x,[Athlete] joins [Competition],x,,,,
|
115 |
+
composer,,,,music by,,,
|
116 |
+
presented by,x,[Award] is presented by [Company],,,,,
|
117 |
+
conflict,,,,,,x,
|
118 |
+
connected with,,,,,,x,
|
119 |
+
conservation status,,,,,,x,
|
120 |
+
central bank,x,[Bank] is the central bank of [Country],x,,,,
|
121 |
+
carries,x,[Bridge] crosses [Artifact],,,,,
|
122 |
+
bridge over,x,[Bridge] crosses [River],,,,,
|
123 |
+
architectural style,x,[Building] has an architectural style of [Person],,,,,
|
124 |
+
contains,,,,is in,,,
|
125 |
+
constructed from,x,[Building] is built from [Material],,,,,
|
126 |
+
cathedral,x,[Cathedral] is in [City],x,,,,
|
127 |
+
costume designer,,,,,work,,
|
128 |
+
ore,x,[Chemical Material] is made from [Stone],,,,,
|
129 |
+
country,,,,,made from,,
|
130 |
+
country of origin,,,,,,x,
|
131 |
+
capital town,x,[City] is a capital town of [Country],x,,,,
|
132 |
+
county seat of,,,,,county seat,,
|
133 |
+
twin city,x,[City] is a twin city of [City],,,,,
|
134 |
+
craft,,,,job,,,
|
135 |
+
city,x,[City] is in [Country],,,,,
|
136 |
+
created by,,,,,work,,
|
137 |
+
created from,,,,formed from,,,
|
138 |
+
creator,,,,,work,,
|
139 |
+
crosses,,,,,carries,,
|
140 |
+
capital of,x,[City] is the capital of [Country],,,,,
|
141 |
+
HQ,x,[Company] has the headquarter in [Location],,,,,
|
142 |
+
divisions,x,[Company] is a subsidiary of [Company],x,,,,
|
143 |
+
sector,x,[Company] is in a sector of [Sector],,,,,
|
144 |
+
dad,,,,father,,,
|
145 |
+
daughter,,,,,father,,
|
146 |
+
daughters,,,,,father,,
|
147 |
+
dead,,,,died on,,,
|
148 |
+
death,,,,died on,,,
|
149 |
+
operated by,x,[Company] operates [Vehicle],,,,,
|
150 |
+
product,x,[Company] owns [Product],,,,,
|
151 |
+
publisher,x,[Company] publishes [Art Work],,,,,
|
152 |
+
depicts,,,,painting of,,,
|
153 |
+
derived from,,,,,,x,
|
154 |
+
designation,,,x,,,,
|
155 |
+
designed by,,,,,work,,
|
156 |
+
designer,,,,,work,,
|
157 |
+
sport,x,[Competition] is a league of [Sport],,,,,
|
158 |
+
developer,,,,developed by,,,
|
159 |
+
topic,x,[Conference] is about [Topic],,,,,
|
160 |
+
council,x,[Council] is the council of [Country],x,,,,
|
161 |
+
claimed by,x,[Country] claims [City],x,,,,
|
162 |
+
diet,,,,,,x,
|
163 |
+
diocese,,,,,,x,
|
164 |
+
diplomatic relation,,,,,,x,
|
165 |
+
directed by,,,,director,,,
|
166 |
+
history,x,[Country] has [History],,,,,
|
167 |
+
director of photography,,,,director,,,
|
168 |
+
assembly,x,[Country] is [Political Party] assembly,,,,,
|
169 |
+
disciple,,,,,student of,,
|
170 |
+
disciple of,,,,student of,,,
|
171 |
+
disciples,,,,,student of,,
|
172 |
+
member of,x,[Country] is a member of [Group],,,,,
|
173 |
+
discography,,,,,,x,
|
174 |
+
discovered by,,,,discoverer,,,
|
175 |
+
enclave within,x,[Country] is enclaved by [Country],,,,,
|
176 |
+
disestablished,,,,,,x,
|
177 |
+
continent,x,[Country] is in [Continent],,,,,
|
178 |
+
distinct from,,,,,,x,
|
179 |
+
distributed by,,,,film studio,,,
|
180 |
+
distribution,,,,,,x,
|
181 |
+
distributor,,,,film studio,,,
|
182 |
+
divided into,,,,,city,,
|
183 |
+
division,,,,sports league,,,
|
184 |
+
athlete,x,[Country] joins [War],x,,,,
|
185 |
+
county seat,x,[Country]'s county seat is [Location],,,,,
|
186 |
+
flag,x,[Country]'s flag is [Artifact],,,,,
|
187 |
+
culture,x,[Culture] is originated in [Country],x,,,,
|
188 |
+
edited by,,,,author,,,
|
189 |
+
currency,x,[Currency] is used in [Country],x,,,,
|
190 |
+
editions,,,,,,x,
|
191 |
+
editor,,,,author,,,
|
192 |
+
educated at,,,,studied at,,,
|
193 |
+
education,,,,studied at,,,
|
194 |
+
effect,,,,,,x,
|
195 |
+
symptoms,x,[Disease] has [Symptoms],,,,,
|
196 |
+
employed by,,,,works at,,,
|
197 |
+
employment,,,,job,,,
|
198 |
+
caused by,x,[Disease] is caused by [Virus],,,,,
|
199 |
+
precedes,x,[Era] comes before [Era],,,,,
|
200 |
+
endemic to,,,,,,x,
|
201 |
+
ending,x,[Event] ends on [Date],,,,,
|
202 |
+
time,x,[Event] happens on [Date],,,,,
|
203 |
+
epoch,,,,era,,,
|
204 |
+
since,x,[Event] is since [Date],,,,,
|
205 |
+
starting,x,[Event] starts on [Date],,,,,
|
206 |
+
ethnic group,,,,,,x,
|
207 |
+
ethnicity,,,,nationality,,,
|
208 |
+
event,,,,took part,,,
|
209 |
+
took part,x,[Event] takes place at [Location],x,,,,
|
210 |
+
executive body,,,,government,,,
|
211 |
+
executive branch,,,,government,,,
|
212 |
+
executive producer,,,,film studio,,,
|
213 |
+
faith,,,,religion,,,
|
214 |
+
family,,,,,,x,
|
215 |
+
famous works,,,,work,,,
|
216 |
+
mascot,x,[Fictional Character] is a mascot of [Sport Team],x,,,,
|
217 |
+
fields,,,,discipline,,,
|
218 |
+
film director,,,,director,,,
|
219 |
+
film editor,,,,author,,,
|
220 |
+
characters,x,[Fictional Character] is from [Art Work] ,x,,,,
|
221 |
+
film producer,,,,film studio,,,
|
222 |
+
made from,x,[Food] is made from [Ingredient],,,,,
|
223 |
+
cuisine,x,[Food] is originated in [Country],,,,,
|
224 |
+
fruit of,x,[Fruit] is [Type],,,,,
|
225 |
+
video game publisher,x,[Game] is produced by [Company],,,,,
|
226 |
+
first language,,,,native language,,,
|
227 |
+
deity of,x,[God] is the god of [Religion],,,,,
|
228 |
+
government,x,[Government] is the government of [Country],x,,,,
|
229 |
+
jurisdiction,x,[Government] is the jurisdiction of [City],,,,,
|
230 |
+
followed by,,,,,series,,
|
231 |
+
follows,,,,series,,,
|
232 |
+
fork of,,,,,,x,
|
233 |
+
form of government,,,,government,,,
|
234 |
+
section of,x,[Group] has a section of [Group],,,,,
|
235 |
+
religion,x,[Group] is [Religion],,,,,
|
236 |
+
formed out of,,,,,,x,
|
237 |
+
successor,x,[Group] is a predecessor of [Group],,,,,
|
238 |
+
founded,,,,foundation,,,
|
239 |
+
founded by,,,,founder,,,
|
240 |
+
religious order,x,[Group] is a religious order of [Group],,,,,
|
241 |
+
created,x,[Group] is created on [Date],,,,,
|
242 |
+
founded in,x,[Group] is founded at [Location],,,,,
|
243 |
+
founder,x,[Group] is founded by [Person],,,,,
|
244 |
+
game engine,,,,video game publisher,,,
|
245 |
+
gameplay,,,,,,x,
|
246 |
+
foundation,x,[Group] is founded on [Date],,,,,
|
247 |
+
gauge,,,,,,x,
|
248 |
+
general manager,,,,team manager,,,
|
249 |
+
funded by,x,[Group] is funded by [Person],,,,,
|
250 |
+
girlfriend,,,,husband,,,
|
251 |
+
given by,,,,presented by,,,
|
252 |
+
legislative body,x,[Group] is legislative body of [Country],x,,,,
|
253 |
+
graduate of,,,,studied at,,,
|
254 |
+
graduated from,,,,studied at,,,
|
255 |
+
ground,,,,home field,,,
|
256 |
+
group,,,,music by,,,
|
257 |
+
happens in,,,,,,x,
|
258 |
+
legislature,x,[Group] is legislature of [Country],x,,,,
|
259 |
+
merged into,x,[Group] is merged into [Group],,,,,
|
260 |
+
parliament,x,[Group] is the parliament of [Country],x,,,,
|
261 |
+
heir,,,,,,x,
|
262 |
+
official language,x,[Group] speaks [Language],,,,,
|
263 |
+
highest point,,,,highest peak,,,
|
264 |
+
highway system,,,,,,x,
|
265 |
+
leader,x,[Group]'s leader is [Person],,,,,
|
266 |
+
appointed by,x,[Head of Government] is appointed by [Head of Government],,,,,
|
267 |
+
home ground,,,,home field,,,
|
268 |
+
home venue,,,,home field,,,
|
269 |
+
honors,,,,award,,,
|
270 |
+
honours,,,,award,,,
|
271 |
+
public holiday,x,[Holiday] is a national holiday of [Country],x,,,,
|
272 |
+
island,x,[Island] is [Country],,,,,
|
273 |
+
head of state,x,[Job] is the head of state in [Location],x,,,,
|
274 |
+
land,x,[Land] is [Country],,,,,
|
275 |
+
alphabet,x,[Language] consists of [Alphabet],,,,,
|
276 |
+
illustrated by,,,,,work,,
|
277 |
+
illustrator,,,,,work,,
|
278 |
+
imprisoned in,,,,prison,,,
|
279 |
+
inaugurated,,,,established,,,
|
280 |
+
include,,,,,is in,,
|
281 |
+
includes,,,,,is in,,
|
282 |
+
includes part,,,,,,x,
|
283 |
+
incorporated,,,,established,,,
|
284 |
+
incumbent,,,,,,x,
|
285 |
+
dialect,x,[Language] is a dialect of [Language],x,,,,
|
286 |
+
indigenous to,,,,culture,,,
|
287 |
+
industry,,,,sector,,,
|
288 |
+
inflow,,,,,,x,
|
289 |
+
enacted by,x,[Law] is enacted by [Group],,,,,
|
290 |
+
informed by,,,,,,x,
|
291 |
+
ingredient,,,,made from,,,
|
292 |
+
inspiration,,,,influenced by,,,
|
293 |
+
inspired by,,,,influenced by,,,
|
294 |
+
instance of,,,,is a,,,
|
295 |
+
instruction set,,,,,,x,
|
296 |
+
instrument,,,,plays,,,
|
297 |
+
intersex,,,,male,,,
|
298 |
+
invented by,,,,,work,,
|
299 |
+
inventor,,,,,work,,
|
300 |
+
is a,,,,,,,x
|
301 |
+
is a type of,,,,is a,,,
|
302 |
+
is an,,,,is a,,,
|
303 |
+
is in,,,,,,,x
|
304 |
+
is in the borough of,,,,town,,,
|
305 |
+
is in the city of,,,,city,,,
|
306 |
+
is in the county of,,,,town,,,
|
307 |
+
is in the district of,,,,town,,,
|
308 |
+
is in the municipality of,,,,town,,,
|
309 |
+
is in the parish of,,,,,,x,
|
310 |
+
is in the province of,,,,state,,,
|
311 |
+
is in the region of,,,,is in,,,
|
312 |
+
is in the state of,,,,state,,,
|
313 |
+
is in the town of,,,,state,,,
|
314 |
+
is in the village of,,,,town,,,
|
315 |
+
is located in,,,,is in,,,
|
316 |
+
is not,,,,,,x,
|
317 |
+
is on,,,,,,x,
|
318 |
+
is owned by,,,,,product,,
|
319 |
+
ISA,,,,,,x,
|
320 |
+
approved by,x,[License] is approved by [Organization],,,,,
|
321 |
+
ballpark,x,[Location] is a ballpark of [Sport Team],x,,,,
|
322 |
+
river mouth,x,[Location] is a river mouth of [Bay],,,,,
|
323 |
+
kids,,,,,father,,
|
324 |
+
killed by,,,,murdered by,,,
|
325 |
+
killer,,,,murdered by,,,
|
326 |
+
kind of,,,,is a,,,
|
327 |
+
sovereign state,x,[Location] is a sovereign state of [Location],,,,,
|
328 |
+
known for,,,,work,,,
|
329 |
+
label,,,,record label,,,
|
330 |
+
administrative centre,x,[Location] is an administrative center of [Location],,,,,
|
331 |
+
language,,,,native language,,,
|
332 |
+
language official,,,,native language,,,
|
333 |
+
language spoken,,,,native language,,,
|
334 |
+
latitude,,,,is in,,,
|
335 |
+
launch date,,,,established,,,
|
336 |
+
launch vehicle,,,,,,x,
|
337 |
+
Indian reservation,x,[Location] is an Indian reservation in [Country],,,,,
|
338 |
+
league,,,,sports league,,,
|
339 |
+
closed,x,[Location] is closed on [Date],,,,,
|
340 |
+
exclave of,x,[Location] is exclave of [Country],,,,,
|
341 |
+
librettist,,,,libretto by,,,
|
342 |
+
planet,x,[Location] is in [Planet],,,,,
|
343 |
+
licence,,,,license,,,
|
344 |
+
next to,x,[Location] is next to [Location],,,,,
|
345 |
+
life partner,,,,,husband,,
|
346 |
+
on the coast of,x,[Location] is on the coast of [Ocean],,,,,
|
347 |
+
literary works,,,,work,,,
|
348 |
+
lived in,,,,residence,,,
|
349 |
+
locality,,,,city,,,
|
350 |
+
located in,,,,is in,,,
|
351 |
+
location,,,,,,x,
|
352 |
+
longitude,,,,,,x,
|
353 |
+
lover,,,,husband,,,
|
354 |
+
split from,x,[Location] is split from [Location],,,,,
|
355 |
+
lyricist,,,,music by,,,
|
356 |
+
lyrics by,,,,music by,,,
|
357 |
+
made by,,,,,product,,
|
358 |
+
highest peak,x,[Location] is the highest peak in [Country],x,,,,
|
359 |
+
made of,,,,made from,,,
|
360 |
+
lowest point,x,[Location] is the lowest point of [Location],x,,,,
|
361 |
+
crystal system,x,[Material]'s crystal system is [Crystal System],,,,,
|
362 |
+
maintenance,,,,maintained by,,,
|
363 |
+
major works,,,,work,,,
|
364 |
+
maker,,,,,product,,
|
365 |
+
makes,,,,product,,,
|
366 |
+
treats,x,[Medication] is for [Disease],,,,,
|
367 |
+
man,,,,male,,,
|
368 |
+
managed by,,,,maintained by,,,
|
369 |
+
manufactured by,,,,manufacturer,,,
|
370 |
+
used for treatment,x,[Medicine] is for [Disease],x,,,,
|
371 |
+
manufactures,,,,,manufacturer,,
|
372 |
+
married to,,,,marry,,,
|
373 |
+
film genre,x,[Movie] is [Genre],,,,,
|
374 |
+
libretto by,x,[Movie] is a libretto by [Person],,,,,
|
375 |
+
master,,,,student of,,,
|
376 |
+
spin-off,x,[Movie] is a spinoff of [Movie],x,,,,
|
377 |
+
medals,,,,,,x,
|
378 |
+
medium,,,,made from,,,
|
379 |
+
member,,,,composed of,,,
|
380 |
+
first broadcast,x,[Movie] is first broadcast on [Date],,,,,
|
381 |
+
members,,,,composed of,,,
|
382 |
+
universe,x,[Movie] is in the universe of [Art Work],,,,,
|
383 |
+
mentor,,,,student of,,,
|
384 |
+
film studio,x,[Movie] is produced by [Company],,,,,
|
385 |
+
merged with,,,,merged into,,,
|
386 |
+
screenplay by,x,[Movie] is screenplayed by [Person],,,,,
|
387 |
+
mode,,,,,,x,
|
388 |
+
film starring,x,[Movie] stars [Actor],,,,,
|
389 |
+
month,,,,,,x,
|
390 |
+
mother,,,,father,,,
|
391 |
+
mother tongue,,,,native language,,,
|
392 |
+
collection,x,[Museum] has [Collection of Work],x,,,,
|
393 |
+
dedicated to,x,[Museum] is dedicated to [Person],,,,,
|
394 |
+
movie director,,,,director,,,
|
395 |
+
municipal council,,,,council,,,
|
396 |
+
memorial to,x,[Museum] is memorial to [Event],,,,,
|
397 |
+
murderer,,,,murdered by,,,
|
398 |
+
genre,x,[Music Artist] is [Genre],,,,,
|
399 |
+
music genre,,,,genre,,,
|
400 |
+
music group,,,,music by,,,
|
401 |
+
musical artist,,,,music by,,,
|
402 |
+
musical score by,,,,music by,,,
|
403 |
+
musician,,,,music by,,,
|
404 |
+
anthem,x,[Music] is an anthem of [Country],x,,,,
|
405 |
+
music by,x,[Music] is made by [Artist],,,,,
|
406 |
+
studio,x,[Music] is recorded at [Studio],,,,,
|
407 |
+
nation,,,,city,,,
|
408 |
+
national holiday,,,,public holiday,,,
|
409 |
+
released,x,[Music] is released on [Date],,,,,
|
410 |
+
official residence,x,[Occupation] lives in [Location],,,,,
|
411 |
+
native to,,,,culture,,,
|
412 |
+
network,,,,broadcast on,,,
|
413 |
+
ideology,x,[Organization]'s ideology is [Ideology],,,,,
|
414 |
+
noble family,,,,,,x,
|
415 |
+
nominated for,,,,award,,,
|
416 |
+
nominee for,,,,award,,,
|
417 |
+
not to be confused with,,,,,,x,
|
418 |
+
notable work,,,,,author,,
|
419 |
+
occupation,,,,job,,,
|
420 |
+
occupied by,,,,maintained by,,,
|
421 |
+
of team,,,,played for,,,
|
422 |
+
office held,,,,,,x,
|
423 |
+
CPU,x,[PC]'s cpu is [CPU],,,,,
|
424 |
+
marry,x,[Person] and [Person] are married,,,,,
|
425 |
+
political party,x,[Person] belongs to [Political Party],,,,,
|
426 |
+
on the shore of,,,,on the coast of,,,
|
427 |
+
opened,,,,starting,,,
|
428 |
+
operated,,,,operated by,,,
|
429 |
+
record label,x,[Person] belongs to [Record Label],,,,,
|
430 |
+
operates,,,,operated by,,,
|
431 |
+
builder,x,[Person] built [Artifact],x,,,,
|
432 |
+
operator,,,,operated by,,,
|
433 |
+
war,x,[Person] causes [War],,,,,
|
434 |
+
orbits,,,,orbit,,,
|
435 |
+
choreographer,x,[Person] choreographs [Play],x,,,,
|
436 |
+
organizer,,,,maintained by,,,
|
437 |
+
original language,,,,native language,,,
|
438 |
+
originates from,,,,,,x,
|
439 |
+
OS,,,,operating system,,,
|
440 |
+
coined by,x,[Person] coins [Artifact],x,,,,
|
441 |
+
outlet,,,,outflow,,,
|
442 |
+
overlies,,,,,,x,
|
443 |
+
owned by,,,,,divisions,,
|
444 |
+
owner,,,,,divisions,,
|
445 |
+
owner of,,,,divisions,,,
|
446 |
+
work,x,[Person] creates [Work],,,,,
|
447 |
+
died in,x,[Person] dies at [Location],,,,,
|
448 |
+
parent,,,,father,,,
|
449 |
+
parent company,,,,,divisions,,
|
450 |
+
parent company of,,,,divisions,,,
|
451 |
+
died on,x,[Person] dies on [Date],,,,,
|
452 |
+
part of,,,,is in,,,
|
453 |
+
participant,,,,player,,,
|
454 |
+
participant in,,,,,player,,
|
455 |
+
participant of,,,,,player,,
|
456 |
+
partner,,,,,husband,,
|
457 |
+
parts,,,,is in,,,
|
458 |
+
party,,,,,,x,
|
459 |
+
passed by,,,,signatory,,,
|
460 |
+
disappeared,x,[Person] disappears on [Date],,,,,
|
461 |
+
first described by,x,[Person] first finds [Disease],x,,,,
|
462 |
+
patron saint of,,,,,patron saint,,
|
463 |
+
pendant of,,,,,,x,
|
464 |
+
disorder,x,[Person] has [Diseases],,,,,
|
465 |
+
period,,,,era,,,
|
466 |
+
house,x,[Person] has a house in [Location],,,,,
|
467 |
+
place of birth,,,,birth place,,,
|
468 |
+
place of death,,,,died in,,,
|
469 |
+
job,x,[Person] is [Occupation],,,,,
|
470 |
+
rank,x,[Person] is [Rank],,,,,
|
471 |
+
platforms,,,,platform,,,
|
472 |
+
male,x,[Person] is [Sex],,,,,
|
473 |
+
candidate,x,[Person] is a candidate of [Election],x,,,,
|
474 |
+
chancellor,x,[Person] is a chancellor of [Country],x,,,,
|
475 |
+
chief executive,x,[Person] is a chief executive of [Company],x,,,,
|
476 |
+
plays for,,,,played for,,,
|
477 |
+
father,x,[Person] is a child of [Person],,,,,
|
478 |
+
portrait of,,,,painting of,,,
|
479 |
+
portrayed by,,,,,work,,
|
480 |
+
cinematographer,x,[Person] is a cinematographer of [Movie],x,,,,
|
481 |
+
predecessor,,,,,successor,,
|
482 |
+
head coach,x,[Person] is a coach of [Sport Team],x,,,,
|
483 |
+
premiere,,,,first performance,,,
|
484 |
+
commander of,x,[Person] is a commander of [Army],,,,,
|
485 |
+
concubine,x,[Person] is a concubine of [Person],x,,,,
|
486 |
+
consort,x,[Person] is a consort of [Person],,,,,
|
487 |
+
husband,x,[Person] is a husband of [Person],,,,,
|
488 |
+
team manager,x,[Person] is a manager of [Sport Team],x,,,,
|
489 |
+
composed of,x,[Person] is a member of [Music Group],x,,,,
|
490 |
+
produces,,,,producer,,,
|
491 |
+
mistress,x,[Person] is a mistress of [Person],x,,,,
|
492 |
+
production company,,,,film studio,,,
|
493 |
+
production designer,,,,,work,,
|
494 |
+
production house,,,,film studio,,,
|
495 |
+
profession,,,,job,,,
|
496 |
+
professor,,,,student of,,,
|
497 |
+
programmer,,,,,,x,
|
498 |
+
patron,x,[Person] is a patron of [Person],,,,,
|
499 |
+
promotion to,,,,,,x,
|
500 |
+
protection,,,,,,x,
|
501 |
+
protocol,,,,,,x,
|
502 |
+
premier,x,[Person] is a premier of [Group],x,,,,
|
503 |
+
public office,,,,,,x,
|
504 |
+
presenter,x,[Person] is a presenter of [TV show],x,,,,
|
505 |
+
student of,x,[Person] is a student of [Person],,,,,
|
506 |
+
active in,x,[Person] is active in [Location],,,,,
|
507 |
+
publishing house,,,,publisher,,,
|
508 |
+
pupil,,,,student of,,,
|
509 |
+
pupil of,,,,student of,,,
|
510 |
+
pupils,,,,student of,,,
|
511 |
+
purpose,,,,use,,,
|
512 |
+
award,x,[Person] is awarded by [Award],,,,,
|
513 |
+
race,,,,nationality,,,
|
514 |
+
nationality,x,[Person] is born in [Country],,,,,
|
515 |
+
radio station,,,,radio network,,,
|
516 |
+
birth place,x,[Person] is born in [Location],,,,,
|
517 |
+
birth date,x,[Person] is born on [Date],,,,,
|
518 |
+
tomb,x,[Person] is buried at [Tomb],,,,,
|
519 |
+
record producer,,,,producer,,,
|
520 |
+
regulatory body,,,,,,x,
|
521 |
+
relative,,,,,,x,
|
522 |
+
convicted of,x,[Person] is convicted of [Crime],,,,,
|
523 |
+
relegation to,,,,,,x,
|
524 |
+
drafted by,x,[Person] is drafted by [Group],,,,,
|
525 |
+
era,x,[Person] is from [Era],,,,,
|
526 |
+
replaced,,,,,successor,,
|
527 |
+
replaced by,,,,,successor,,
|
528 |
+
representative body,,,,,,x,
|
529 |
+
represented by,,,,head of government,,,
|
530 |
+
prison,x,[Person] is in [Prison],,,,,
|
531 |
+
resident in,,,,residence,,,
|
532 |
+
resting place,,,,tomb,,,
|
533 |
+
murdered by,x,[Person] is killed by [Person],,,,,
|
534 |
+
result of,,,,,result,,
|
535 |
+
played for,x,[Person] is played at [Group],,,,,
|
536 |
+
royal house,,,,dynasty,,,
|
537 |
+
runs on,,,,,,x,
|
538 |
+
chairman,x,[Person] is the chair of [Group],x,,,,
|
539 |
+
school,,,,studied at,,,
|
540 |
+
emperor,x,[Person] is the emperor of [Country],x,,,,
|
541 |
+
screenwriter,,,,screenplay by,,,
|
542 |
+
script,,,,alphabet,,,
|
543 |
+
scriptwriter,,,,,,x,
|
544 |
+
head of government,x,[Person] is the head of government of [Country],x,,,,
|
545 |
+
seat,,,,county seat,,,
|
546 |
+
king,x,[Person] is the king of [Country],x,,,,
|
547 |
+
dynasty,x,[Person] is the leader of [Dynasty],,,,,
|
548 |
+
senior coach,,,,head coach,,,
|
549 |
+
separated from,,,,,,x,
|
550 |
+
sequel of,,,,series,,,
|
551 |
+
mayor,x,[Person] is the mayor of [City],x,,,,
|
552 |
+
set of,,,,,,x,
|
553 |
+
sex,,,,male,,,
|
554 |
+
monarch,x,[Person] is the monarch of [Country],x,,,,
|
555 |
+
shareholder,,,,,,x,
|
556 |
+
shares border with,,,,next to,,,
|
557 |
+
president,x,[Person] is the president of [Country],x,,,,
|
558 |
+
signed by,,,,signatory,,,
|
559 |
+
significant works,,,,work,,,
|
560 |
+
prime minister,x,[Person] is the prime minister of [Country],x,,,,
|
561 |
+
singer,,,,music by,,,
|
562 |
+
sister city,,,,twin city,,,
|
563 |
+
solid solution series with,,,,,,x,
|
564 |
+
son,,,,,father,,
|
565 |
+
songwriter,,,,author,,,
|
566 |
+
sons,,,,,father,,
|
567 |
+
queen,x,[Person] is the queen of [Country] ,x,,,,
|
568 |
+
residence,x,[Person] live in [Location],,,,,
|
569 |
+
played by,x,[Person] plays [Fictional Character],x,,,,
|
570 |
+
plays,x,[Person] plays [Instrument],,,,,
|
571 |
+
sports,,,,sport,,,
|
572 |
+
voice actor,x,[Person] plays in [Movie],,,,,
|
573 |
+
spouse,,,,,husband,,
|
574 |
+
square,,,,is in,,,
|
575 |
+
stadium,,,,home field,,,
|
576 |
+
starring,,,,film starring,,,
|
577 |
+
team,x,[Person] plays in [Sport Team],,,,,
|
578 |
+
native language,x,[Person] speaks [Language],,,,,
|
579 |
+
stock exchange,,,,,,x,
|
580 |
+
discipline,x,[Person] studies [Academic Subject],,,,,
|
581 |
+
student,,,,,student of,,
|
582 |
+
studied at,x,[Person] studies at [School],,,,,
|
583 |
+
students,,,,,student of,,
|
584 |
+
works at,x,[Person] works at [Group],,,,,
|
585 |
+
studied under,,,,student of,,,
|
586 |
+
pet,x,[Pet] is a pet of [Person],x,,,,
|
587 |
+
satellite,x,[Planet] is a satellite of [Planet],x,,,,
|
588 |
+
subclass of,,,,is a,,,
|
589 |
+
subdivided into,,,,,city,,
|
590 |
+
subject,,,,,,x,
|
591 |
+
subject of,,,,,,x,
|
592 |
+
subsidiary,,,,divisions,,,
|
593 |
+
subsidiary company,,,,divisions,,,
|
594 |
+
subsidiary of,,,,divisions,,,
|
595 |
+
subsystem of,,,,,,x,
|
596 |
+
orbit,x,[Planet] is in the orbit of [Orbit],,,,,
|
597 |
+
honours,,,,studied at,,,
|
598 |
+
type species,x,[Plant] is [Type],,,,,
|
599 |
+
first performance,x,[Play] is first performed on [Date],,,,,
|
600 |
+
takes place in,,,,,,x,
|
601 |
+
taxonomic rank,,,,,,x,
|
602 |
+
teacher,,,,student of,,,
|
603 |
+
teacher of,,,,,student of,,
|
604 |
+
performer,x,[Play] is performed by [Person],,,,,
|
605 |
+
radio network,x,[Radio Program] is broadcasted on [Radio Channel],,,,,
|
606 |
+
teams played for,,,,played for,,,
|
607 |
+
teleplay by,,,,,,x,
|
608 |
+
television channel,,,,broadcast on,,,
|
609 |
+
tenant,,,,,,x,
|
610 |
+
railway line,x,[Railway] is in [Location],x,,,,
|
611 |
+
territory claimed by,,,,claimed by,,,
|
612 |
+
theme song,,,,music by,,,
|
613 |
+
denomination,x,[Religion] is a denomination by [Artifact],,,,,
|
614 |
+
drain,x,[River] drains [Location],,,,,
|
615 |
+
tributary of,x,[River] is a tributary of [River],,,,,
|
616 |
+
tonality,,,,,,x,
|
617 |
+
outflow,x,[River] outflows to [Location],,,,,
|
618 |
+
took part in,,,,took part,,,
|
619 |
+
developed by,x,[Software] is developed by [Company],,,,,
|
620 |
+
license,x,[Software] is under license of [License],,,,,
|
621 |
+
use,x,[Software] is used for [Purpose],,,,,
|
622 |
+
treatment,,,,treats,,,
|
623 |
+
programming language,x,[Software] is written in [Programming Language],,,,,
|
624 |
+
tributary,,,,,tributary of,,
|
625 |
+
affiliate of,x,[Sport Team] is an affiliate of [Sport Team],,,,,
|
626 |
+
tunnel under,,,,bridge over,,,
|
627 |
+
TV channel,,,,broadcast on,,,
|
628 |
+
sports league,x,[Sport Team] plays at [Competition],,,,,
|
629 |
+
player,x,[Sport Team] plays in [Competition],x,,,,
|
630 |
+
champion,x,[Sport Team] wins [Competition],x,,,,
|
631 |
+
unit of,,,,,,x,
|
632 |
+
home field,x,[Sport Team]'s home field is [Location],,,,,
|
633 |
+
constellation,x,[Star] is a [Constellation],,,,,
|
634 |
+
state,x,[State] is a state of [Country],,,,,
|
635 |
+
used for,,,,use,,,
|
636 |
+
terminus,x,[Station] is the terminus of [Railway],x,,,,
|
637 |
+
used in,,,,use,,,
|
638 |
+
user,,,,,,x,
|
639 |
+
uses,,,,,use,,
|
640 |
+
utility,,,,use,,,
|
641 |
+
venue,,,,,took part,,
|
642 |
+
street,x,[Street] is in the street of [Location],,,,,
|
643 |
+
system of,x,[System] is a system in [Artifact],,,,,
|
644 |
+
voiced by,,,,voice actor,,,
|
645 |
+
time zone,x,[Timezone] is a timezon in [Country],x,,,,
|
646 |
+
town,x,[Town] is in [Location],,,,,
|
647 |
+
wife,,,,,husband,,
|
648 |
+
winner,,,,champion,,,
|
649 |
+
winners,,,,champion,,,
|
650 |
+
woman,,,,male,,,
|
651 |
+
won by,,,,champion,,,
|
652 |
+
words by,,,,author,,,
|
653 |
+
broadcast on,x,[TV Series] is broadcasted on [TV Channel],,,,,
|
654 |
+
working for,,,,works at,,,
|
655 |
+
works,,,,work,,,
|
656 |
+
weapon,x,[Weapon] is used by [Person],x,,,,
|
657 |
+
writer,,,,author,,,
|
658 |
+
writing system,,,,alphabet,,,
|
659 |
+
written by,,,,author,,,
|
660 |
+
year,,,,time,,,
|
process.py
CHANGED
@@ -1,33 +1,87 @@
|
|
1 |
-
"""
|
|
|
|
|
2 |
wget https://figshare.com/ndownloader/files/8760241
|
3 |
unzip 8760241
|
|
|
4 |
"""
|
5 |
|
6 |
import json
|
|
|
|
|
7 |
import os
|
8 |
from glob import glob
|
9 |
from tqdm import tqdm
|
10 |
-
|
11 |
-
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
|
16 |
-
|
17 |
-
for
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
30 |
data = [json.loads(i) for i in f.read().split('\n') if len(i) > 0]
|
31 |
-
|
32 |
-
|
33 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
""" Process raw T-Rex file.
|
2 |
+
mkdir data_raw
|
3 |
+
cd data_raw
|
4 |
wget https://figshare.com/ndownloader/files/8760241
|
5 |
unzip 8760241
|
6 |
+
cd ../
|
7 |
"""
|
8 |
|
9 |
import json
|
10 |
+
import string
|
11 |
+
import re
|
12 |
import os
|
13 |
from glob import glob
|
14 |
from tqdm import tqdm
|
15 |
+
|
16 |
+
import pandas as pd
|
17 |
+
|
18 |
+
# process raw data
|
19 |
+
if not os.path.exists('data/t_rex.raw.jsonl'):
|
20 |
+
os.makedirs('data', exist_ok=True)
|
21 |
+
f_writer = open('data/t_rex.raw.jsonl', 'w')
|
22 |
+
for i in tqdm(glob("data_raw/*.json")):
|
23 |
+
with open(i) as f:
|
24 |
+
data = json.load(f)
|
25 |
+
for _data in data:
|
26 |
+
for triple in _data['triples']:
|
27 |
+
p = triple['predicate']['surfaceform']
|
28 |
+
o = triple['object']['surfaceform']
|
29 |
+
s = triple['subject']['surfaceform']
|
30 |
+
if p is None or o is None or s is None:
|
31 |
+
continue
|
32 |
+
out = {"predicate": p, "object": o, "subject": s, "title": _data["title"], "text": _data["text"]}
|
33 |
+
f_writer.write(json.dumps(out) + "\n")
|
34 |
+
f_writer.close()
|
35 |
+
|
36 |
+
# apply filtering to remove noisy instances
|
37 |
+
stopwords = ["he", "she", "they", "it"]
|
38 |
+
list_alnum = string.ascii_lowercase + '0123456789 '
|
39 |
+
|
40 |
+
|
41 |
+
def filtering(entry):
|
42 |
+
|
43 |
+
def _subfilter(token):
|
44 |
+
if len(re.findall(rf'[^{list_alnum}]+', token)) != 0:
|
45 |
+
return False
|
46 |
+
if token in stopwords:
|
47 |
+
return False
|
48 |
+
if token.startswith("www"):
|
49 |
+
return False
|
50 |
+
if token.startswith("."):
|
51 |
+
return False
|
52 |
+
if token.startswith(","):
|
53 |
+
return False
|
54 |
+
if token.startswith("$"):
|
55 |
+
return False
|
56 |
+
if token.startswith("+"):
|
57 |
+
return False
|
58 |
+
if token.startswith("#"):
|
59 |
+
return False
|
60 |
+
return True
|
61 |
+
|
62 |
+
if not _subfilter(entry["object"].lower()):
|
63 |
+
return False
|
64 |
+
if not _subfilter(entry["subject"].lower()):
|
65 |
+
return False
|
66 |
+
|
67 |
+
if entry['object'].islower() and entry['subject'].islower():
|
68 |
+
return False
|
69 |
+
|
70 |
+
return True
|
71 |
+
|
72 |
+
|
73 |
+
with open(f"data/t_rex.raw.jsonl") as f:
|
74 |
data = [json.loads(i) for i in f.read().split('\n') if len(i) > 0]
|
75 |
+
print(f"[before]: {len(data)}")
|
76 |
+
data = [i for i in data if filtering(i)]
|
77 |
+
df = pd.DataFrame(data)
|
78 |
+
count = df.groupby("predicate")['title'].count()
|
79 |
+
df = df[[count[p] >= 3 for p in df['predicate']]]
|
80 |
+
df = df.drop_duplicates()
|
81 |
+
print(f"[after] : {len(df)}")
|
82 |
+
print(f"[entity]: {len(set(df['object'].unique().tolist() + df['subject'].unique().tolist()))}")
|
83 |
+
_df = df[['object', 'subject']]
|
84 |
+
data = [i.to_dict() for _, i in df.iterrows()]
|
85 |
+
|
86 |
+
with open(f"data/t_rex.filter.jsonl", 'w') as f:
|
87 |
+
f.write('\n'.join([json.dumps(i) for i in data]))
|
t_rex.py
CHANGED
@@ -5,7 +5,7 @@ import datasets
|
|
5 |
logger = datasets.logging.get_logger(__name__)
|
6 |
_DESCRIPTION = """T-Rex dataset."""
|
7 |
_NAME = "t_rex"
|
8 |
-
_VERSION = "0.0.
|
9 |
_CITATION = """
|
10 |
@inproceedings{elsahar2018t,
|
11 |
title={T-rex: A large scale alignment of natural language with knowledge base triples},
|
@@ -17,8 +17,9 @@ _CITATION = """
|
|
17 |
|
18 |
_HOME_PAGE = "https://github.com/asahi417/relbert"
|
19 |
_URL = f'https://huggingface.co/datasets/relbert/{_NAME}/resolve/main/data'
|
20 |
-
_TYPES = ["raw", "filter"]
|
21 |
-
|
|
|
22 |
str(datasets.Split.TRAIN): [f'{_URL}/t_rex.{i}.train.jsonl'],
|
23 |
str(datasets.Split.VALIDATION): [f'{_URL}/t_rex.{i}.validation.jsonl'],
|
24 |
str(datasets.Split.TEST): [f'{_URL}/t_rex.{i}.test.jsonl']}
|
@@ -46,7 +47,7 @@ class TREX(datasets.GeneratorBasedBuilder):
|
|
46 |
|
47 |
def _split_generators(self, dl_manager):
|
48 |
downloaded_file = dl_manager.download_and_extract(_URLS[self.config.name])
|
49 |
-
if self.config.name in
|
50 |
return [datasets.SplitGenerator(
|
51 |
name=datasets.Split.TRAIN,
|
52 |
gen_kwargs={"filepaths": downloaded_file[str(datasets.Split.TRAIN)]})]
|
|
|
5 |
logger = datasets.logging.get_logger(__name__)
|
6 |
_DESCRIPTION = """T-Rex dataset."""
|
7 |
_NAME = "t_rex"
|
8 |
+
_VERSION = "0.0.4"
|
9 |
_CITATION = """
|
10 |
@inproceedings{elsahar2018t,
|
11 |
title={T-rex: A large scale alignment of natural language with knowledge base triples},
|
|
|
17 |
|
18 |
_HOME_PAGE = "https://github.com/asahi417/relbert"
|
19 |
_URL = f'https://huggingface.co/datasets/relbert/{_NAME}/resolve/main/data'
|
20 |
+
_TYPES = ["raw", "filter", "filter_unified"]
|
21 |
+
_NON_SPLITS = ["raw", "filter", "filter_unified"]
|
22 |
+
_URLS = {i: {str(datasets.Split.TRAIN): [f'{_URL}/t_rex.{i}.jsonl']} if i in _NON_SPLITS else {
|
23 |
str(datasets.Split.TRAIN): [f'{_URL}/t_rex.{i}.train.jsonl'],
|
24 |
str(datasets.Split.VALIDATION): [f'{_URL}/t_rex.{i}.validation.jsonl'],
|
25 |
str(datasets.Split.TEST): [f'{_URL}/t_rex.{i}.test.jsonl']}
|
|
|
47 |
|
48 |
def _split_generators(self, dl_manager):
|
49 |
downloaded_file = dl_manager.download_and_extract(_URLS[self.config.name])
|
50 |
+
if self.config.name in _NON_SPLITS:
|
51 |
return [datasets.SplitGenerator(
|
52 |
name=datasets.Split.TRAIN,
|
53 |
gen_kwargs={"filepaths": downloaded_file[str(datasets.Split.TRAIN)]})]
|
unify_predicate.py
ADDED
@@ -0,0 +1,69 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# import json
|
2 |
+
# import pandas as pd
|
3 |
+
#
|
4 |
+
# with open("data/t_rex.filter.jsonl") as f:
|
5 |
+
# data = pd.DataFrame([json.loads(i) for i in f.read().split('\n') if len(i) > 0])
|
6 |
+
# freq = data.groupby("predicate").count()['title']
|
7 |
+
# data['freq'] = [freq.loc[i] for i in data['predicate']]
|
8 |
+
# data = data[data['freq'] >= 3]
|
9 |
+
# tmp = data.groupby("predicate").sample(10, replace=True)
|
10 |
+
# tmp = tmp.drop_duplicates()
|
11 |
+
# tmp.to_csv("data/t_rex.filter.predicate_check_sample.csv", index=False)
|
12 |
+
|
13 |
+
import json
|
14 |
+
import pandas as pd
|
15 |
+
|
16 |
+
# load manual check sheet
|
17 |
+
df_predicate = pd.read_csv('predicate_manual_check.csv')
|
18 |
+
df_predicate = df_predicate[df_predicate['remove (noisy)'] != 'x']
|
19 |
+
df_predicate = df_predicate[df_predicate['remove (too vague)'] != 'x']
|
20 |
+
predicate_main = df_predicate[df_predicate['ok'] == 'x']['unique predicates'].tolist()
|
21 |
+
df_sub = df_predicate[df_predicate['ok'] != 'x']
|
22 |
+
df_sub_same = df_sub[['unique predicates', 'same as']].dropna()
|
23 |
+
df_sub_same = df_sub_same[[i in predicate_main for i in df_sub_same['same as']]]
|
24 |
+
df_sub_same.index = df_sub_same.pop('unique predicates')
|
25 |
+
sub_same = df_sub_same['same as'].to_dict()
|
26 |
+
df_sub_rev = df_sub[['unique predicates', 'reverse of']].dropna()
|
27 |
+
df_sub_rev = df_sub_rev[[i in predicate_main for i in df_sub_rev['reverse of']]]
|
28 |
+
df_sub_rev.index = df_sub_rev.pop('unique predicates')
|
29 |
+
sub_rev = df_sub_rev['reverse of'].to_dict()
|
30 |
+
|
31 |
+
# load data and filter based on manual predicate check sheet
|
32 |
+
with open(f"data/t_rex.filter.jsonl") as f:
|
33 |
+
data = pd.DataFrame([json.loads(i) for i in f.read().split('\n') if len(i) > 0])
|
34 |
+
data['predicate'] = [sub_same[i] if i in sub_same else i for i in data['predicate']]
|
35 |
+
data['reverse'] = [i in sub_rev for i in data['predicate']]
|
36 |
+
data['predicate'] = [sub_rev[i] if i in sub_rev else i for i in data['predicate']]
|
37 |
+
data_filter = data[[i in predicate_main for i in data['predicate']]]
|
38 |
+
|
39 |
+
data_filter_rev = data_filter[data_filter['reverse']].copy()
|
40 |
+
o = data_filter_rev.pop("object")
|
41 |
+
s = data_filter_rev.pop("subject")
|
42 |
+
data_filter_rev["subject"] = o
|
43 |
+
data_filter_rev["object"] = s
|
44 |
+
data_filter[data_filter['reverse']] = data_filter_rev[data_filter.columns]
|
45 |
+
data_filter.pop("reverse")
|
46 |
+
|
47 |
+
df_main = df_predicate[df_predicate['ok'] == 'x'][['unique predicates', 'pretty relation name', 'pretty relation name is reverse']]
|
48 |
+
df_main['reverse'] = [i == 'x' for i in df_main.pop('pretty relation name is reverse')]
|
49 |
+
df_main['predicate'] = df_main.pop('unique predicates')
|
50 |
+
|
51 |
+
data_filter_join = data_filter.merge(df_main, how='inner', on='predicate')
|
52 |
+
data_filter_join_rev = data_filter_join[data_filter_join['reverse']].copy()
|
53 |
+
o = data_filter_join_rev.pop("object")
|
54 |
+
s = data_filter_join_rev.pop("subject")
|
55 |
+
data_filter_join_rev["subject"] = o
|
56 |
+
data_filter_join_rev["object"] = s
|
57 |
+
data_filter_join[data_filter_join['reverse']] = data_filter_join_rev[data_filter_join.columns]
|
58 |
+
data_filter_join.pop("reverse")
|
59 |
+
data_filter_join.pop("predicate")
|
60 |
+
data_filter_join['predicate'] = data_filter_join.pop("pretty relation name")
|
61 |
+
|
62 |
+
print(f"[after] : {len(data_filter_join)}")
|
63 |
+
print(f"[entity]: {len(set(data_filter_join['object'].unique().tolist() + data_filter_join['subject'].unique().tolist()))}")
|
64 |
+
print(f"[predicate]: {len(data_filter_join['predicate'].unique())}")
|
65 |
+
|
66 |
+
|
67 |
+
data = [i.to_dict() for _, i in data_filter_join.iterrows()]
|
68 |
+
with open(f"data/t_rex.filter_unified.jsonl", 'w') as f:
|
69 |
+
f.write('\n'.join([json.dumps(i) for i in data]))
|