Datasets:

Languages:
English
ArXiv:
ChancesYuan commited on
Commit
b2de267
·
1 Parent(s): 3ad4879

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +8 -8
README.md CHANGED
@@ -27,13 +27,14 @@ We build four datasets for the sub-task of EDIT and ADD based on two benchmark d
27
  ### Data Instances
28
 
29
  An example of E-FB15k237:
 
30
 
31
  ```
32
  {
33
- "ori": ,
34
- "cor": ,
35
- "process": ,
36
- "label":
37
  }
38
  ```
39
 
@@ -41,10 +42,9 @@ An example of A-FB15k237:
41
 
42
  ```
43
  {
44
- "ori": ,
45
- "cor": ,
46
- "process": ,
47
- "label":
48
  }
49
  ```
50
 
 
27
  ### Data Instances
28
 
29
  An example of E-FB15k237:
30
+ (Note that we have converted the ID to text for easier understanding)
31
 
32
  ```
33
  {
34
+ "ori": ["Jennifer Connelly", "type of union", "Marriage"],
35
+ "cor": ["Stephen Sondheim", "type of union", "Marriage"],
36
+ "process": ["[MASK]", "type of union", "Marriage"],
37
+ "label": "Jennifer Connelly"
38
  }
39
  ```
40
 
 
42
 
43
  ```
44
  {
45
+ "triples": ["Darryl F. Zanuck", "place of death", "Palm Springs"],
46
+ "label": "Palm Springs",
47
+ "head": 0
 
48
  }
49
  ```
50