Datasets:
Sub-tasks:
named-entity-recognition
Multilinguality:
multilingual
Size Categories:
unknown
Language Creators:
machine-generated
Annotations Creators:
expert-generated
Source Datasets:
original
ArXiv:
Tags:
word-segmentation
License:
Commit
•
0c78895
1
Parent(s):
088c9bb
Fix `license` metadata (#1)
Browse files- Fix `license` metadata (e3454351ea8da55eda4b8bed6639e560d3824b5b)
Co-authored-by: Julien Chaumond <julien-c@users.noreply.huggingface.co>
README.md
CHANGED
|
@@ -1,148 +1,148 @@
|
|
| 1 |
-
---
|
| 2 |
-
annotations_creators:
|
| 3 |
-
- expert-generated
|
| 4 |
-
language_creators:
|
| 5 |
-
- machine-generated
|
| 6 |
-
|
| 7 |
-
- hi
|
| 8 |
-
- en
|
| 9 |
-
|
| 10 |
-
- unknown
|
| 11 |
-
multilinguality:
|
| 12 |
-
- multilingual
|
| 13 |
-
pretty_name: HashSet Manual
|
| 14 |
-
size_categories:
|
| 15 |
-
- unknown
|
| 16 |
-
source_datasets:
|
| 17 |
-
- original
|
| 18 |
-
task_categories:
|
| 19 |
-
- structure-prediction
|
| 20 |
-
task_ids:
|
| 21 |
-
- named-entity-recognition
|
| 22 |
-
- structure-prediction-other-word-segmentation
|
| 23 |
-
---
|
| 24 |
-
|
| 25 |
-
# Dataset Card for HashSet Manual
|
| 26 |
-
|
| 27 |
-
## Dataset Description
|
| 28 |
-
|
| 29 |
-
- **Repository:** [prashantkodali/HashSet](https://github.com/prashantkodali/HashSet)
|
| 30 |
-
- **Paper:** [HashSet -- A Dataset For Hashtag Segmentation](https://arxiv.org/abs/2201.06741)
|
| 31 |
-
|
| 32 |
-
### Dataset Summary
|
| 33 |
-
|
| 34 |
-
Hashset is a new dataset consisting on 1.9k manually annotated and 3.3M loosely supervised tweets for testing the
|
| 35 |
-
efficiency of hashtag segmentation models. We compare State of The Art Hashtag Segmentation models on Hashset and other
|
| 36 |
-
baseline datasets (STAN and BOUN). We compare and analyse the results across the datasets to argue that HashSet can act
|
| 37 |
-
as a good benchmark for hashtag segmentation tasks.
|
| 38 |
-
|
| 39 |
-
HashSet Manual: contains 1.9k manually annotated hashtags. Each row consists of the hashtag, segmented hashtag ,named entity annotations, whether the hashtag contains mix of hindi and english tokens and/or contains non-english tokens.
|
| 40 |
-
|
| 41 |
-
### Languages
|
| 42 |
-
|
| 43 |
-
Mostly Hindi and English.
|
| 44 |
-
|
| 45 |
-
## Dataset Structure
|
| 46 |
-
|
| 47 |
-
### Data Instances
|
| 48 |
-
|
| 49 |
-
```
|
| 50 |
-
{
|
| 51 |
-
"index": 10,
|
| 52 |
-
"hashtag": "goodnewsmegan",
|
| 53 |
-
"segmentation": "good news megan",
|
| 54 |
-
"spans": {
|
| 55 |
-
"start": [
|
| 56 |
-
8
|
| 57 |
-
],
|
| 58 |
-
"end": [
|
| 59 |
-
13
|
| 60 |
-
],
|
| 61 |
-
"text": [
|
| 62 |
-
"megan"
|
| 63 |
-
]
|
| 64 |
-
},
|
| 65 |
-
"source": "roman",
|
| 66 |
-
"gold_position": null,
|
| 67 |
-
"mix": false,
|
| 68 |
-
"other": false,
|
| 69 |
-
"ner": true,
|
| 70 |
-
"annotator_id": 1,
|
| 71 |
-
"annotation_id": 2088,
|
| 72 |
-
"created_at": "2021-12-30 17:10:33.800607",
|
| 73 |
-
"updated_at": "2021-12-30 17:10:59.714840",
|
| 74 |
-
"lead_time": 3896.182,
|
| 75 |
-
"rank": {
|
| 76 |
-
"position": [
|
| 77 |
-
1,
|
| 78 |
-
2,
|
| 79 |
-
3,
|
| 80 |
-
4,
|
| 81 |
-
5,
|
| 82 |
-
6,
|
| 83 |
-
7,
|
| 84 |
-
8,
|
| 85 |
-
9,
|
| 86 |
-
10
|
| 87 |
-
],
|
| 88 |
-
"candidate": [
|
| 89 |
-
"goodnewsmegan",
|
| 90 |
-
"goodnewsmeg an",
|
| 91 |
-
"goodnews megan",
|
| 92 |
-
"goodnewsmega n",
|
| 93 |
-
"go odnewsmegan",
|
| 94 |
-
"good news megan",
|
| 95 |
-
"good newsmegan",
|
| 96 |
-
"g oodnewsmegan",
|
| 97 |
-
"goodnewsme gan",
|
| 98 |
-
"goodnewsm egan"
|
| 99 |
-
]
|
| 100 |
-
}
|
| 101 |
-
}
|
| 102 |
-
```
|
| 103 |
-
|
| 104 |
-
### Data Fields
|
| 105 |
-
|
| 106 |
-
- `index`: a numerical index annotated by Kodali et al..
|
| 107 |
-
- `hashtag`: the original hashtag.
|
| 108 |
-
- `segmentation`: the gold segmentation for the hashtag.
|
| 109 |
-
- `spans`: named entity spans.
|
| 110 |
-
- `source`: data source.
|
| 111 |
-
- `gold_position`: position of the gold segmentation on the `segmentation` field inside the `rank`.
|
| 112 |
-
- `mix`: The hashtag has a mix of English and Hindi tokens.
|
| 113 |
-
- `other`: The hashtag has non-English tokens.
|
| 114 |
-
- `ner`: The hashtag has named entities.
|
| 115 |
-
- `annotator_id`: annotator ID.
|
| 116 |
-
- `annotation_id`: annotation ID.
|
| 117 |
-
- `created_at`: Creation date timestamp.
|
| 118 |
-
- `updated_at`: Update date timestamp.
|
| 119 |
-
- `lead_time`: Lead time field annotated by Kodali et al..
|
| 120 |
-
- `rank`: Rank of each candidate selected by a baseline word segmenter ( WordBreaker ).
|
| 121 |
-
- `candidates`: Candidates selected by a baseline word segmenter ( WordBreaker ).
|
| 122 |
-
|
| 123 |
-
## Dataset Creation
|
| 124 |
-
|
| 125 |
-
- All hashtag segmentation and identifier splitting datasets on this profile have the same basic fields: `hashtag` and `segmentation` or `identifier` and `segmentation`.
|
| 126 |
-
|
| 127 |
-
- The only difference between `hashtag` and `segmentation` or between `identifier` and `segmentation` are the whitespace characters. Spell checking, expanding abbreviations or correcting characters to uppercase go into other fields.
|
| 128 |
-
|
| 129 |
-
- There is always whitespace between an alphanumeric character and a sequence of any special characters ( such as `_` , `:`, `~` ).
|
| 130 |
-
|
| 131 |
-
- If there are any annotations for named entity recognition and other token classification tasks, they are given in a `spans` field.
|
| 132 |
-
|
| 133 |
-
## Additional Information
|
| 134 |
-
|
| 135 |
-
### Citation Information
|
| 136 |
-
|
| 137 |
-
```
|
| 138 |
-
@article{kodali2022hashset,
|
| 139 |
-
title={HashSet--A Dataset For Hashtag Segmentation},
|
| 140 |
-
author={Kodali, Prashant and Bhatnagar, Akshala and Ahuja, Naman and Shrivastava, Manish and Kumaraguru, Ponnurangam},
|
| 141 |
-
journal={arXiv preprint arXiv:2201.06741},
|
| 142 |
-
year={2022}
|
| 143 |
-
}
|
| 144 |
-
```
|
| 145 |
-
|
| 146 |
-
### Contributions
|
| 147 |
-
|
| 148 |
This dataset was added by [@ruanchaves](https://github.com/ruanchaves) while developing the [hashformers](https://github.com/ruanchaves/hashformers) library.
|
|
|
|
| 1 |
+
---
|
| 2 |
+
annotations_creators:
|
| 3 |
+
- expert-generated
|
| 4 |
+
language_creators:
|
| 5 |
+
- machine-generated
|
| 6 |
+
language:
|
| 7 |
+
- hi
|
| 8 |
+
- en
|
| 9 |
+
license:
|
| 10 |
+
- unknown
|
| 11 |
+
multilinguality:
|
| 12 |
+
- multilingual
|
| 13 |
+
pretty_name: HashSet Manual
|
| 14 |
+
size_categories:
|
| 15 |
+
- unknown
|
| 16 |
+
source_datasets:
|
| 17 |
+
- original
|
| 18 |
+
task_categories:
|
| 19 |
+
- structure-prediction
|
| 20 |
+
task_ids:
|
| 21 |
+
- named-entity-recognition
|
| 22 |
+
- structure-prediction-other-word-segmentation
|
| 23 |
+
---
|
| 24 |
+
|
| 25 |
+
# Dataset Card for HashSet Manual
|
| 26 |
+
|
| 27 |
+
## Dataset Description
|
| 28 |
+
|
| 29 |
+
- **Repository:** [prashantkodali/HashSet](https://github.com/prashantkodali/HashSet)
|
| 30 |
+
- **Paper:** [HashSet -- A Dataset For Hashtag Segmentation](https://arxiv.org/abs/2201.06741)
|
| 31 |
+
|
| 32 |
+
### Dataset Summary
|
| 33 |
+
|
| 34 |
+
Hashset is a new dataset consisting on 1.9k manually annotated and 3.3M loosely supervised tweets for testing the
|
| 35 |
+
efficiency of hashtag segmentation models. We compare State of The Art Hashtag Segmentation models on Hashset and other
|
| 36 |
+
baseline datasets (STAN and BOUN). We compare and analyse the results across the datasets to argue that HashSet can act
|
| 37 |
+
as a good benchmark for hashtag segmentation tasks.
|
| 38 |
+
|
| 39 |
+
HashSet Manual: contains 1.9k manually annotated hashtags. Each row consists of the hashtag, segmented hashtag ,named entity annotations, whether the hashtag contains mix of hindi and english tokens and/or contains non-english tokens.
|
| 40 |
+
|
| 41 |
+
### Languages
|
| 42 |
+
|
| 43 |
+
Mostly Hindi and English.
|
| 44 |
+
|
| 45 |
+
## Dataset Structure
|
| 46 |
+
|
| 47 |
+
### Data Instances
|
| 48 |
+
|
| 49 |
+
```
|
| 50 |
+
{
|
| 51 |
+
"index": 10,
|
| 52 |
+
"hashtag": "goodnewsmegan",
|
| 53 |
+
"segmentation": "good news megan",
|
| 54 |
+
"spans": {
|
| 55 |
+
"start": [
|
| 56 |
+
8
|
| 57 |
+
],
|
| 58 |
+
"end": [
|
| 59 |
+
13
|
| 60 |
+
],
|
| 61 |
+
"text": [
|
| 62 |
+
"megan"
|
| 63 |
+
]
|
| 64 |
+
},
|
| 65 |
+
"source": "roman",
|
| 66 |
+
"gold_position": null,
|
| 67 |
+
"mix": false,
|
| 68 |
+
"other": false,
|
| 69 |
+
"ner": true,
|
| 70 |
+
"annotator_id": 1,
|
| 71 |
+
"annotation_id": 2088,
|
| 72 |
+
"created_at": "2021-12-30 17:10:33.800607",
|
| 73 |
+
"updated_at": "2021-12-30 17:10:59.714840",
|
| 74 |
+
"lead_time": 3896.182,
|
| 75 |
+
"rank": {
|
| 76 |
+
"position": [
|
| 77 |
+
1,
|
| 78 |
+
2,
|
| 79 |
+
3,
|
| 80 |
+
4,
|
| 81 |
+
5,
|
| 82 |
+
6,
|
| 83 |
+
7,
|
| 84 |
+
8,
|
| 85 |
+
9,
|
| 86 |
+
10
|
| 87 |
+
],
|
| 88 |
+
"candidate": [
|
| 89 |
+
"goodnewsmegan",
|
| 90 |
+
"goodnewsmeg an",
|
| 91 |
+
"goodnews megan",
|
| 92 |
+
"goodnewsmega n",
|
| 93 |
+
"go odnewsmegan",
|
| 94 |
+
"good news megan",
|
| 95 |
+
"good newsmegan",
|
| 96 |
+
"g oodnewsmegan",
|
| 97 |
+
"goodnewsme gan",
|
| 98 |
+
"goodnewsm egan"
|
| 99 |
+
]
|
| 100 |
+
}
|
| 101 |
+
}
|
| 102 |
+
```
|
| 103 |
+
|
| 104 |
+
### Data Fields
|
| 105 |
+
|
| 106 |
+
- `index`: a numerical index annotated by Kodali et al..
|
| 107 |
+
- `hashtag`: the original hashtag.
|
| 108 |
+
- `segmentation`: the gold segmentation for the hashtag.
|
| 109 |
+
- `spans`: named entity spans.
|
| 110 |
+
- `source`: data source.
|
| 111 |
+
- `gold_position`: position of the gold segmentation on the `segmentation` field inside the `rank`.
|
| 112 |
+
- `mix`: The hashtag has a mix of English and Hindi tokens.
|
| 113 |
+
- `other`: The hashtag has non-English tokens.
|
| 114 |
+
- `ner`: The hashtag has named entities.
|
| 115 |
+
- `annotator_id`: annotator ID.
|
| 116 |
+
- `annotation_id`: annotation ID.
|
| 117 |
+
- `created_at`: Creation date timestamp.
|
| 118 |
+
- `updated_at`: Update date timestamp.
|
| 119 |
+
- `lead_time`: Lead time field annotated by Kodali et al..
|
| 120 |
+
- `rank`: Rank of each candidate selected by a baseline word segmenter ( WordBreaker ).
|
| 121 |
+
- `candidates`: Candidates selected by a baseline word segmenter ( WordBreaker ).
|
| 122 |
+
|
| 123 |
+
## Dataset Creation
|
| 124 |
+
|
| 125 |
+
- All hashtag segmentation and identifier splitting datasets on this profile have the same basic fields: `hashtag` and `segmentation` or `identifier` and `segmentation`.
|
| 126 |
+
|
| 127 |
+
- The only difference between `hashtag` and `segmentation` or between `identifier` and `segmentation` are the whitespace characters. Spell checking, expanding abbreviations or correcting characters to uppercase go into other fields.
|
| 128 |
+
|
| 129 |
+
- There is always whitespace between an alphanumeric character and a sequence of any special characters ( such as `_` , `:`, `~` ).
|
| 130 |
+
|
| 131 |
+
- If there are any annotations for named entity recognition and other token classification tasks, they are given in a `spans` field.
|
| 132 |
+
|
| 133 |
+
## Additional Information
|
| 134 |
+
|
| 135 |
+
### Citation Information
|
| 136 |
+
|
| 137 |
+
```
|
| 138 |
+
@article{kodali2022hashset,
|
| 139 |
+
title={HashSet--A Dataset For Hashtag Segmentation},
|
| 140 |
+
author={Kodali, Prashant and Bhatnagar, Akshala and Ahuja, Naman and Shrivastava, Manish and Kumaraguru, Ponnurangam},
|
| 141 |
+
journal={arXiv preprint arXiv:2201.06741},
|
| 142 |
+
year={2022}
|
| 143 |
+
}
|
| 144 |
+
```
|
| 145 |
+
|
| 146 |
+
### Contributions
|
| 147 |
+
|
| 148 |
This dataset was added by [@ruanchaves](https://github.com/ruanchaves) while developing the [hashformers](https://github.com/ruanchaves/hashformers) library.
|