Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
10K - 100K
License:
updated
Browse files
README.md
CHANGED
|
@@ -50,6 +50,9 @@ The HumAID Twitter dataset consists of several thousands of manually annotated t
|
|
| 50 |
### Supported Tasks and Benchmark
|
| 51 |
The dataset can be used to train a model for multiclass tweet classification for disaster response. The benchmark results can be found in https://ojs.aaai.org/index.php/ICWSM/article/view/18116/17919.
|
| 52 |
|
|
|
|
|
|
|
|
|
|
| 53 |
### Languages
|
| 54 |
|
| 55 |
English
|
|
@@ -114,6 +117,7 @@ AMT has been used to annotate the dataset. Please check the paper for a more det
|
|
| 114 |
|
| 115 |
Authors of the paper.
|
| 116 |
|
|
|
|
| 117 |
### Licensing Information
|
| 118 |
|
| 119 |
- cc-by-nc-4.0
|
|
|
|
| 50 |
### Supported Tasks and Benchmark
|
| 51 |
The dataset can be used to train a model for multiclass tweet classification for disaster response. The benchmark results can be found in https://ojs.aaai.org/index.php/ICWSM/article/view/18116/17919.
|
| 52 |
|
| 53 |
+
Dataset is also released with event-wise and JSON objects for further research.
|
| 54 |
+
Full set of the dataset can be found in https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/A7NVF7
|
| 55 |
+
|
| 56 |
### Languages
|
| 57 |
|
| 58 |
English
|
|
|
|
| 117 |
|
| 118 |
Authors of the paper.
|
| 119 |
|
| 120 |
+
|
| 121 |
### Licensing Information
|
| 122 |
|
| 123 |
- cc-by-nc-4.0
|