chnln commited on
Commit
72f58e2
·
verified ·
1 Parent(s): 6fc33a2

Release SINS binary common-ground judgment dataset

Browse files
Files changed (2) hide show
  1. README.md +46 -0
  2. data/train-00000-of-00001.parquet +3 -0
README.md ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-4.0
3
+ task_categories:
4
+ - text-classification
5
+ language:
6
+ - en
7
+ tags:
8
+ - dialogue
9
+ - common-ground
10
+ - maptask
11
+ - vision-language
12
+ size_categories:
13
+ - 10K<n<100K
14
+ ---
15
+
16
+ # Seeing Is Not Sharing (SINS) Binary Common-Ground Judgment Dataset
17
+
18
+ ## Dataset Description
19
+
20
+ SINS is a 13,077-instance binary common-ground judgment dataset derived from
21
+ the public Grounded Misunderstandings in MapTask (GMMT) annotations. Each row
22
+ preserves identifiers and context-window provenance while mapping GMMT status
23
+ to `gold_label`: `aligned` becomes `yes`; `pending` and `misunderstood` become
24
+ `no`.
25
+
26
+ ## Data Fields
27
+
28
+ The release contains `ref_id`, `dialogue_id`, `map_id`, `utt_id`,
29
+ `transaction_id`, `context_transaction_ids`, `end_utt_id_of_context`,
30
+ `timed_unit_ids`, `expression`, `status`, and `gold_label`.
31
+
32
+ ## Excluded Material
33
+
34
+ The SINS HF dataset does not contain dialogue context. It does not contain MapTask maps,
35
+ images, OCR, or image-derived text. It also does not contain
36
+ generated prompts, model predictions, log-probabilities, or copies of the
37
+ underlying MapTask source files. Users who have authorised local access to
38
+ MapTask may reconstruct context with this repository's scripts and a local
39
+ GMMT checkout; reconstructed files must remain local.
40
+
41
+ ## Licensing and Provenance
42
+
43
+ SINS is released under CC BY 4.0. It is derived from
44
+ [`chnln/grounded-misunderstandings-in-maptask`](https://huggingface.co/datasets/chnln/grounded-misunderstandings-in-maptask),
45
+ which is also CC BY 4.0. Please cite both the SINS paper and GMMT when using
46
+ this dataset. See `NOTICE` in the code repository for details.
data/train-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:327f6285218877134b60e674d134e8797f96848c2b5ec48060b488dcce05f64b
3
+ size 409543