Datasets:
boyuanzheng010
commited on
Commit
•
305a151
1
Parent(s):
c406d5a
Update README.md
Browse files
README.md
CHANGED
@@ -60,6 +60,16 @@ Multimodal-Mind2Web is the multimodal version of [Mind2Web](https://osu-nlp-grou
|
|
60 |
|
61 |
## Dataset Structure
|
62 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
63 |
|
64 |
### Data Fields
|
65 |
|
@@ -69,30 +79,22 @@ Multimodal-Mind2Web is the multimodal version of [Mind2Web](https://osu-nlp-grou
|
|
69 |
- "subdomain" (str): website subdomain
|
70 |
- "confirmed_task" (str): task description
|
71 |
- "action_reprs" (list[str]): human readable string representation of the action sequence
|
72 |
-
- "
|
73 |
-
-
|
74 |
-
|
75 |
-
|
76 |
-
|
77 |
-
|
78 |
-
|
79 |
-
|
80 |
-
|
81 |
-
|
82 |
-
|
83 |
-
|
84 |
-
|
85 |
-
|
86 |
-
|
87 |
-
- "neg_candidates" (list[dict]): other candidate elements in the page after preprocessing, has similar structure as "pos_candidates"
|
88 |
|
89 |
-
### Data Splits
|
90 |
-
|
91 |
-
- train: 1,009 instances
|
92 |
-
- test:
|
93 |
-
- Cross Task: 177 instances, tasks from the same website are seen during training
|
94 |
-
- Cross Website: 142 instances, websites are not seen during training
|
95 |
-
- Cross Domain: 694 instances, entire domains are not seen during training
|
96 |
|
97 |
|
98 |
### Disclaimer
|
|
|
60 |
|
61 |
## Dataset Structure
|
62 |
|
63 |
+
### Data Splits
|
64 |
+
- train: 1009 tasks.
|
65 |
+
- test:
|
66 |
+
- Cross Task: 1339 actions from 177 tasks. Tasks from the same website are seen during training.
|
67 |
+
- Cross Website: 1019 actions from 142 tasks. Websites are not seen during training.
|
68 |
+
- Cross Domain: 4060 actions from 694 tasks. Entire domains are not seen during training.
|
69 |
+
|
70 |
+
The three test set splits have went through human verification to confirm element visibility and correct rendering for action prediction.
|
71 |
+
|
72 |
+
Each line in the dataset is an action consisting of screenshot image, HTML text and other fields required for action prediction, for the convenience of inference.
|
73 |
|
74 |
### Data Fields
|
75 |
|
|
|
79 |
- "subdomain" (str): website subdomain
|
80 |
- "confirmed_task" (str): task description
|
81 |
- "action_reprs" (list[str]): human readable string representation of the action sequence
|
82 |
+
- "screenshot" (str): path to the webpage screenshot image corresponding to the HTML.
|
83 |
+
- "action_uid" (str): unique id for each action (step)
|
84 |
+
- "raw_html" (str): raw html of the page before the action is performed
|
85 |
+
- "cleaned_html" (str): cleaned html of the page before the action is performed
|
86 |
+
- "operation" (dict): operation to perform
|
87 |
+
- "op" (str): operation type, one of CLICK, TYPE, SELECT
|
88 |
+
- "original_op" (str): original operation type, contain additional HOVER and ENTER that are mapped to CLICK, not used
|
89 |
+
- "value" (str): optional value for the operation, e.g., text to type, option to select
|
90 |
+
- "pos_candidates" (list[dict]): ground truth elements. Here we only include positive elements that exist in "cleaned_html" after our preprocessing, so "pos_candidates" might be empty. The original labeled element can always be found in the "raw_html".
|
91 |
+
- "tag" (str): tag of the element
|
92 |
+
- "is_original_target" (bool): whether the element is the original target labeled by the annotator
|
93 |
+
- "is_top_level_target" (bool): whether the element is a top level target find by our algorithm. please see the paper for more details.
|
94 |
+
- "backend_node_id" (str): unique id for the element
|
95 |
+
- "attributes" (str): serialized attributes of the element, use `json.loads` to convert back to dict
|
96 |
+
- "neg_candidates" (list[dict]): other candidate elements in the page after preprocessing, has similar structure as "pos_candidates"
|
|
|
97 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
98 |
|
99 |
|
100 |
### Disclaimer
|