Enhancing Natural Language Inference Performance with Knowledge Graph for COVID-19 Automated Fact-Checking in Indonesian Language
2025-08-29
Summary
The study explores the use of Knowledge Graphs (KG) to enhance Natural Language Inference (NLI) for automated COVID-19 fact-checking in the Indonesian language. The proposed model combines KG data with NLI to verify information more accurately, achieving a maximum accuracy of 86.16% using the XLM-RoBERTa pre-trained language model.
Why This Matters
Accurate information dissemination is critical in combating misinformation, especially during health crises like COVID-19. This research highlights the potential of integrating external knowledge sources, such as KGs, to improve automated fact-checking systems, which can be crucial in ensuring public access to reliable health information in languages with fewer resources, like Indonesian.
How You Can Use This Info
Professionals working in media, public health, or technology can leverage enhanced fact-checking tools to better manage misinformation. Integrating KGs into NLI systems can improve the reliability of information dissemination, presenting an opportunity for organizations to develop more robust automated fact-checking capabilities, especially in underrepresented languages.