ClaimPKG: Vietnam's AI technology helps verify information more accurately

(NLDO) - In the context of fake news and misinformation spreading rapidly on social networks, the need to verify content is becoming increasingly urgent.

To solve this problem, Viettel Data Services and Artificial Intelligence Center (Viettel AI) has researched and developed ClaimPKG - a new information verification method applying artificial intelligence, capable of significantly improving accuracy.

ClaimPKG has just been selected to present at the ACL 2025 International Conference - the world's leading prestigious forum in the field of natural language processing (NLP), which brings together technology experts from Google, Meta, IBM, Amazon... Viettel AI's research has surpassed more than 5,200 other topics, with a selection rate of only 37%.

The need to verify information has become an essential "defensive barrier" for every user in the digital space.

Technically, ClaimPKG combines a knowledge graph (Kgraph) with a large language model (LLM). This method increases verification accuracy by 9% to 12% compared to current methods.

In the field of journalism and media, this technology can help reporters and editors detect fake news and compare news sources quickly and with basis.

In the medical and legal fields - industries that require absolutely accurate information, this technology can act as a "verification assistant", helping to look up and authenticate specialized knowledge.

In addition, ClaimPKG also has the potential to be integrated into chatbots or virtual assistants, helping the system provide more reliable answers, avoiding "vague judgment".

The ACL 2025 conference recently held in Vienna (Austria) is one of the most prestigious and influential events globally in the field of NLP processing. The event brings together thousands of researchers, engineers and technology experts from large organizations such as Google, Meta, Amazon, IBM... to share the latest advances in NLP.

The conference content ranges from language machine learning, machine translation, text summarization to ethics and fairness issues in linguistic AI. The main theme this year is the generalization ability of NLP models – a key factor for AI systems to flexibly adapt to new data and contexts.

Source: https://nld.com.vn/claimpkg-cong-nghe-ai-cua-viet-nam-giup-kiem-chung-thong-tin-chinh-xac-hon-196250805073301463.htm