DariMis: Harm-Aware Modeling for Dari Misinformation Detection on YouTube

1. 1Collect and annotate a dataset of Dari-language YouTube videos with labels for information type and harm level
2. 2Develop a harm-aware modeling approach that takes into account the nuances of the Dari language and the specific harm levels associated with different types of misinformation
3. 3Evaluate the performance of the model on the annotated dataset and refine it as needed
4. 4Apply the model to real-world scenarios to detect and mitigate the spread of harmful misinformation on YouTube

AI engineers and researchers working on natural language processing and misinformation detection can benefit from this study, as it provides a new dataset and approach for detecting harmful misinformation in underrepresented languages

💡 The development of harm-aware modeling approaches for misinformation detection can help mitigate the spread of harmful content in underrepresented languages

🚨 DariMis: a new approach for detecting misinformation in Dari-language YouTube videos 📹