This is the official repository for Hostile Post Detection in Hindi language. This dataset was used in the shared task at Constraint@AAAI collocated with AAAI 2021.

The Dataset paper can be found here: https://arxiv.org/abs/2011.03588

For more details about the shared task please visit: https://constraint-shared-task-2021.github.io/

The competition was hosted at CodaLab, for more details please visit: https://competitions.codalab.org/competitions/26654

The shared task consisted of two sub-tasks:

  i.  Coarse grained Hostility Detection: The task is to predict whether a given post is hostile or not.
  ii. Fine-grained Hostility Detection: The task is to detect one or more hostile dimensions which the post belongs to. The four hostile dimensions are fake, hate, defamation, and offensive

A total of 44 teams participated in the competition. The final leaderboard can be found here

@misc{bhardwaj2020hostility, title={Hostility Detection Dataset in Hindi}, author={Mohit Bhardwaj and Md Shad Akhtar and Asif Ekbal and Amitava Das and Tanmoy Chakraborty}, year={2020}, eprint={2011.03588}, archivePrefix={arXiv}, primaryClass={cs.CL} }