HiBias: A Nine-Faceted Bias Annotation Dataset for Media Bias Detection in Hindi

Kushal Trivedi, Roshni Chakraborty, Subhoshri Pal, Murtuza Shaikh, Rohit Kumawat, Anisha Khairnar, Satwik Duppala, Sai Pranav, Chakali Surendra, Dhruv Sanghvi

Keywords: Media Bias, Indic Languages

Abstract:

Although automated detection, understanding and mitigation of media bias is highly required in Indian context, this requires
curation of trustworthy annotated news articles datasets in Indian languages, such as, Hindi, that covers multiple facets of bias. In this paper,
we introduce the first annotated dataset consisting of 400 unique articles from two leading Indian news media agencies in Hindi language. Our annotations include 9 different types of bias, that ensures exhaustiveness and additionally, include explanations from the annotators. The dataset and replication code are publicly available at 1 and 2.

Submission Number: 9