IndicClaimVer@FIRE-2026
Shared Task on Claim Verification and Evidence Retrieval for Indian Languages
Shared Task on Claim Verification and Evidence Retrieval for Indian Languages
The shared task, IndicClaimVer, co-located with the FIRE 2026 conference, focuses on an end-to-end pipeline for evidence retrieval and claim verification in major Indian languages, particularly English, Hindi, Bengali, and CodeMixed. The task is motivated by the increasing spread of misinformation in multilingual Indian digital ecosystems, where content often appears in regional, low-resource, and code-mixed forms. While significant progress has been made on claim verification systems for high-resource languages such as English, a robust evaluation framework for Indic-language claim verification and evidence retrieval still remains limited.
IndicClaimVer@FIRE-2026 aims to address this gap by benchmarking systems across both controlled and open-domain verification settings. The task is structured into two complementary subtasks:
Subtask 1: Evidence Selection + Veracity Prediction (Closed Evidence Setting)
Subtask 2: End-to-End Claim Verification (Open Evidence Setting)
Please fill out the following Google Form in order to participate in the shared task.
Google Form link: https://forms.gle/bEcbX4aQGH1nDpZA7
Each registered participant must have the following credentials:
A valid team name
Valid email addresses of all participants
Institutional affiliation
Website goes live: June 10, 2026
Registration of Teams: June 10 - 30, 2026
Release of Training and Development data: June 15, 2026