Datasets:
Languages:
English
Multilinguality:
monolingual
Size Categories:
10K<n<100K
Annotations Creators:
expert-generated
License:
Update README.md
Browse files
README.md
CHANGED
@@ -60,11 +60,17 @@ task_ids:
|
|
60 |
|
61 |
### Dataset Summary
|
62 |
|
63 |
-
|
64 |
|
65 |
### Supported Tasks and Leaderboards
|
66 |
|
67 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
68 |
|
69 |
### Languages
|
70 |
|
@@ -119,7 +125,11 @@ A typical instance in the dataset:
|
|
119 |
|
120 |
### Data Fields
|
121 |
|
122 |
-
|
|
|
|
|
|
|
|
|
123 |
|
124 |
### Data Splits
|
125 |
|
|
|
60 |
|
61 |
### Dataset Summary
|
62 |
|
63 |
+
Open PI is the first dataset for tracking state changes in procedural text from arbitrary domains by using an unrestricted (open) vocabulary. Our solution is a new task formulation in which just the text is provided, from which a set of state changes (entity, attribute, before, after) is generated for each step, where the entity, attribute, and values must all be predicted from an open vocabulary.
|
64 |
|
65 |
### Supported Tasks and Leaderboards
|
66 |
|
67 |
+
- `Task 1`: Input -> one Paragraph (e.g., with 5 steps), Output -> Entities that change (challenge: implicit entities, some explicit entities that don’t change)
|
68 |
+
|
69 |
+
- `Task 3`: Input -> Paragraph, Output -> <Attr of entity> that change (challenge: implicit entities, attributes & many combinations)
|
70 |
+
|
71 |
+
- `Task 4`: Input -> Paragraph, entity, Output -> sequence of attr value changes (challenge: implicit attributes)
|
72 |
+
|
73 |
+
- `Task 7`: Input -> Image url for a step, Output -> <Visual Attr of entity> and <Non-visual Attr of entity> that change
|
74 |
|
75 |
### Languages
|
76 |
|
|
|
125 |
|
126 |
### Data Fields
|
127 |
|
128 |
+
The following is an excerpt from the dataset README:
|
129 |
+
|
130 |
+
Within "qas", some fields should be self-explanatory. Listed below is an explanation about the others:
|
131 |
+
|
132 |
+
#### Fields specific to questions:
|
133 |
|
134 |
### Data Splits
|
135 |
|