Saved in:
Bibliographic Details
Main Authors: Tran, Dien X., Nguyen, Nam V., Tran, Thanh T., Hoang, Anh T., Duong, Tai V., Le, Di T., Le, Phuc-Lu
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.00955
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915533170933760
author Tran, Dien X.
Nguyen, Nam V.
Tran, Thanh T.
Hoang, Anh T.
Duong, Tai V.
Le, Di T.
Le, Phuc-Lu
author_facet Tran, Dien X.
Nguyen, Nam V.
Tran, Thanh T.
Hoang, Anh T.
Duong, Tai V.
Le, Di T.
Le, Phuc-Lu
contents The rise of misinformation, exacerbated by Large Language Models (LLMs) like GPT and Gemini, demands robust fact-checking solutions, especially for low-resource languages like Vietnamese. Existing methods struggle with semantic ambiguity, homonyms, and complex linguistic structures, often trading accuracy for efficiency. We introduce SemViQA, a novel Vietnamese fact-checking framework integrating Semantic-based Evidence Retrieval (SER) and Two-step Verdict Classification (TVC). Our approach balances precision and speed, achieving state-of-the-art results with 78.97\% strict accuracy on ISE-DSC01 and 80.82\% on ViWikiFC, securing 1st place in the UIT Data Science Challenge. Additionally, SemViQA Faster improves inference speed 7x while maintaining competitive accuracy. SemViQA sets a new benchmark for Vietnamese fact verification, advancing the fight against misinformation. The source code is available at: https://github.com/DAVID-NGUYEN-S16/SemViQA.
format Preprint
id arxiv_https___arxiv_org_abs_2503_00955
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SemViQA: A Semantic Question Answering System for Vietnamese Information Fact-Checking
Tran, Dien X.
Nguyen, Nam V.
Tran, Thanh T.
Hoang, Anh T.
Duong, Tai V.
Le, Di T.
Le, Phuc-Lu
Computation and Language
Artificial Intelligence
The rise of misinformation, exacerbated by Large Language Models (LLMs) like GPT and Gemini, demands robust fact-checking solutions, especially for low-resource languages like Vietnamese. Existing methods struggle with semantic ambiguity, homonyms, and complex linguistic structures, often trading accuracy for efficiency. We introduce SemViQA, a novel Vietnamese fact-checking framework integrating Semantic-based Evidence Retrieval (SER) and Two-step Verdict Classification (TVC). Our approach balances precision and speed, achieving state-of-the-art results with 78.97\% strict accuracy on ISE-DSC01 and 80.82\% on ViWikiFC, securing 1st place in the UIT Data Science Challenge. Additionally, SemViQA Faster improves inference speed 7x while maintaining competitive accuracy. SemViQA sets a new benchmark for Vietnamese fact verification, advancing the fight against misinformation. The source code is available at: https://github.com/DAVID-NGUYEN-S16/SemViQA.
title SemViQA: A Semantic Question Answering System for Vietnamese Information Fact-Checking
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2503.00955