Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
10K - 100K
License:
Commit History
-
updated data-card
93470f4
-
updated
6ae2656 -
updated readme
472fcc2 -
updated data loader script
084936f -
updated
54eee82
-
updated readme
866c07a -
Upload dataset_infos.json
c321fcd -
Upload data/validation-00000-of-00001.parquet
559c193 -
Upload data/test-00000-of-00001.parquet with git-lfs
c832de4 -
Upload data/train-00000-of-00001.parquet with git-lfs
b568e35 -
upload data loading script
6dd03e9 -
updated dataset info
7019ee9 -
dataset info
616a4f8 -
updated readme
32c566a -
dataset loader script
eae4739
-
README.md
Description of the dataset ========================== The Crisis Benchmarks Dataset consists of data from different data sources [1]): Data format and directories =========================== The data directory contains the following sub-directories. In each directory, there are files for Informativeness and Humanitarian tasks. * data/all_data_en -- all combined english dataset used for the experiments * data/individual_data_en -- consists of data used for the experiments as individual data source such as crisisnlp and crisislex * data/event_aware_en -- all combined english dataset with event tag (fire, earthquake, flood, ...) are tagged * data/data_split_all_lang -- all combined dataset with their train/dev and test splits * data/initial_filtering -- all combined dataset duplicate removed data * data/class_label_mapped -- all combined dataset initial set of dataset where class label mapped Format of the TSV files ----------------------------------------------------…1843be7 -
initial commit
9c15c2b