Update README.md
Browse files
README.md
CHANGED
@@ -24,17 +24,15 @@ CLUTRR contains a large set of semi-synthetic stories involving hypothetical fam
|
|
24 |
|
25 |
Join the CLUTRR community in https://www.cs.mcgill.ca/~ksinha4/clutrr/
|
26 |
## Dataset Structure
|
27 |
-
We show detailed information for all
|
28 |
|
29 |
### configurations:
|
30 |
**id**: a unique series of characters and numbers that identify each instance <br>
|
31 |
**story**: one semi-synthetic story involving hypothetical families<br>
|
32 |
**query**: the target query/relation which contains two names, where the goal is to classify the relation that holds between these two entities<br>
|
33 |
-
**text\_query**: <br>
|
34 |
**target**: correct relation for the query <br>
|
35 |
(the following kin-ship relations are used: son, father, husband, brother, grandson, grandfather, son-in-law, father-in-law, brother-in-law, uncle, nephew, daughter, mother, wife, sister, granddaughter, grandmother, daughter-in-law, mother-in-law, sister-in-law, aunt, niece.)<br>
|
36 |
-
**
|
37 |
-
**clean\_story**: <br>
|
38 |
**proof\_state**: the logical rule of the kinship generation <br>
|
39 |
**f\_comb**: the kinships of the query followed by the logical rule<br>
|
40 |
**task\_name**: the task of the sub-dataset in a form of "task_[num1].[num2]"<br>
|
@@ -47,8 +45,6 @@ The second number [num2] directly indicates the length of clauses for the task t
|
|
47 |
**edge\_types**: similar to the f\_comb, another form of the query's kinships followed by the logical rule <br>
|
48 |
**query\_edge**: the corresponding edge of the target query in the kinship graph<br>
|
49 |
**genders**: genders of names appeared in the story<br>
|
50 |
-
**syn\_story**: <br>
|
51 |
-
**node\_mapping**: <br>
|
52 |
**task\_split**: train,test <br>
|
53 |
|
54 |
*Further explanation of Irrelevant facts, Supporting facts and Disconnected facts can be found in the 3.5 Robust Reasoning section in https://arxiv.org/abs/1908.06177
|
@@ -61,9 +57,7 @@ An example of 'train'in Task 1.2 looks as follows.
|
|
61 |
"id": b2b9752f-d7fa-46a9-83ae-d474184c35b6,
|
62 |
"story": "[Lillian] and her daughter [April] went to visit [Lillian]'s mother [Ashley] last Sunday.",
|
63 |
"query": ('April', 'Ashley'),
|
64 |
-
"text_query": ,
|
65 |
"target": "grandmother",
|
66 |
-
"text_target": ['[Ashley] has a granddaughter called [April] who is her favourite.'],
|
67 |
"clean_story": [Lillian] and her daughter [April] went to visit [Lillian]'s mother [Ashley] last Sunday.,
|
68 |
"proof_state": [{('April', 'grandmother', 'Ashley'): [('April', 'mother', 'Lillian'), ('Lillian', 'mother', 'Ashley')]}],
|
69 |
"f_comb": "mother-mother",
|
@@ -71,9 +65,7 @@ An example of 'train'in Task 1.2 looks as follows.
|
|
71 |
"story_edges": [(0, 1), (1, 2)],
|
72 |
"edge_types": ['mother', 'mother'],
|
73 |
"query_edge": (0, 2),
|
74 |
-
"genders": "April:female,Lillian:female,Ashley:female"
|
75 |
-
"syn_story": ,
|
76 |
-
"node_mapping": {7: 0, 2: 1, 1: 2},
|
77 |
"task_split": trian
|
78 |
}
|
79 |
```
|
|
|
24 |
|
25 |
Join the CLUTRR community in https://www.cs.mcgill.ca/~ksinha4/clutrr/
|
26 |
## Dataset Structure
|
27 |
+
We show detailed information for all 13 configurations of the dataset.
|
28 |
|
29 |
### configurations:
|
30 |
**id**: a unique series of characters and numbers that identify each instance <br>
|
31 |
**story**: one semi-synthetic story involving hypothetical families<br>
|
32 |
**query**: the target query/relation which contains two names, where the goal is to classify the relation that holds between these two entities<br>
|
|
|
33 |
**target**: correct relation for the query <br>
|
34 |
(the following kin-ship relations are used: son, father, husband, brother, grandson, grandfather, son-in-law, father-in-law, brother-in-law, uncle, nephew, daughter, mother, wife, sister, granddaughter, grandmother, daughter-in-law, mother-in-law, sister-in-law, aunt, niece.)<br>
|
35 |
+
**clean\_story**: the story without noise factors<br>
|
|
|
36 |
**proof\_state**: the logical rule of the kinship generation <br>
|
37 |
**f\_comb**: the kinships of the query followed by the logical rule<br>
|
38 |
**task\_name**: the task of the sub-dataset in a form of "task_[num1].[num2]"<br>
|
|
|
45 |
**edge\_types**: similar to the f\_comb, another form of the query's kinships followed by the logical rule <br>
|
46 |
**query\_edge**: the corresponding edge of the target query in the kinship graph<br>
|
47 |
**genders**: genders of names appeared in the story<br>
|
|
|
|
|
48 |
**task\_split**: train,test <br>
|
49 |
|
50 |
*Further explanation of Irrelevant facts, Supporting facts and Disconnected facts can be found in the 3.5 Robust Reasoning section in https://arxiv.org/abs/1908.06177
|
|
|
57 |
"id": b2b9752f-d7fa-46a9-83ae-d474184c35b6,
|
58 |
"story": "[Lillian] and her daughter [April] went to visit [Lillian]'s mother [Ashley] last Sunday.",
|
59 |
"query": ('April', 'Ashley'),
|
|
|
60 |
"target": "grandmother",
|
|
|
61 |
"clean_story": [Lillian] and her daughter [April] went to visit [Lillian]'s mother [Ashley] last Sunday.,
|
62 |
"proof_state": [{('April', 'grandmother', 'Ashley'): [('April', 'mother', 'Lillian'), ('Lillian', 'mother', 'Ashley')]}],
|
63 |
"f_comb": "mother-mother",
|
|
|
65 |
"story_edges": [(0, 1), (1, 2)],
|
66 |
"edge_types": ['mother', 'mother'],
|
67 |
"query_edge": (0, 2),
|
68 |
+
"genders": "April:female,Lillian:female,Ashley:female",
|
|
|
|
|
69 |
"task_split": trian
|
70 |
}
|
71 |
```
|