shawon95/

Bengali-Fake-Review-Dataset

This is a binary dataset used for Bengali fake review detection in the paper "Bengali Fake Reviews: A Benchmark Dataset and Detection System" accepted in Neurocomputing, a journal published by Elsevier.

Annotated by 4 native Bangla speakers with more than 90% trustworthiness score.

Fleiss' Kappa Score: 0.83

Fake - 1339 Non-fake - 7710

	Class wise statistics of BFRD dataset




	

Statistics Fake Non-fake

Total words 1,55,789 9,27,902

Total… See the full description on the dataset page: https://huggingface.co/datasets/shawon95/Bengali-Fake-Review-Dataset.