ShortCheck: Checkworthiness Detection of Multilingual Short-Form Videos
This addresses misinformation detection for human fact-checkers in the challenging context of short-form videos, though it appears incremental as it combines existing methods into a new pipeline.
The authors tackled the problem of detecting checkworthy short-form videos on platforms like TikTok by developing ShortCheck, a modular pipeline that integrates speech transcription, OCR, object detection, and other components, achieving an F1-weighted score over 70% on multilingual datasets.
Short-form video platforms like TikTok present unique challenges for misinformation detection due to their multimodal, dynamic, and noisy content. We present ShortCheck, a modular, inference-only pipeline with a user-friendly interface that automatically identifies checkworthy short-form videos to help human fact-checkers. The system integrates speech transcription, OCR, object and deepfake detection, video-to-text summarization, and claim verification. ShortCheck is validated by evaluating it on two manually annotated datasets with TikTok videos in a multilingual setting. The pipeline achieves promising results with F1-weighted score over 70\%.