--- language: - vi library_name: transformers license: mit pipeline_tag: text-classification tags: - SemViQA - three-class-classification - fact-checking --- # SemViQA-TC: Vietnamese Three-class Classification for Claim Verification ## Model Description 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. **SemViQA-TC** is one of the key components of the **SemViQA** system, designed for **three-class classification** in Vietnamese fact-checking. This model classifies a given claim into one of three categories: **SUPPORTED**, **REFUTED**, or **NOT ENOUGH INFORMATION (NEI)** based on retrieved evidence. To address these challenges, SemViQA integrates: - **Semantic-based Evidence Retrieval (SER)**: Combines **TF-IDF** with a **Question Answering Token Classifier (QATC)** to enhance retrieval precision while reducing inference time. - **Two-step Verdict Classification (TVC)**: Uses hierarchical classification optimized with **Cross-Entropy and Focal Loss**, improving claim verification across three categories: - **Supported** ✅ - **Refuted** ❌ - **Not Enough Information (NEI)** 🤷‍♂️ ### **Model Information** - **Developed by:** [SemViQA Research Team](https://huggingface.co/SemViQA) - **Fine-tuned model:** [InfoXLM](https://huggingface.co/microsoft/infoxlm-large) - **Supported Language:** Vietnamese - **Task:** Three-Class Classification (Fact Verification) - **Dataset:** [ViWikiFC](https://arxiv.org/abs/2405.07615) SemViQA-TC serves as the **first step in the two-step classification process** of the SemViQA system. It initially categorizes claims into three classes: **SUPPORTED, REFUTED, or NEI**. For claims classified as **SUPPORTED** or **REFUTED**, a secondary **binary classification model (SemViQA-BC)** further refines the prediction. This hierarchical classification strategy enhances the accuracy of fact verification. ### **🏆 Achievements** - **1st place** in the **UIT Data Science Challenge** 🏅 - **State-of-the-art** performance on: - **ISE-DSC01** → **78.97% strict accuracy** - **ViWikiFC** → **80.82% strict accuracy** - **SemViQA Faster**: **7x speed improvement** over the standard model 🚀 These results establish **SemViQA** as a **benchmark for Vietnamese fact verification**, advancing efforts to combat misinformation and ensure **information integrity**. ## Usage Example Direct Model Usage ```Python # Install semviqa !pip install semviqa # Initalize a pipeline import torch import torch.nn.functional as F from transformers import AutoTokenizer from semviqa.tvc.model import ClaimModelForClassification tokenizer = AutoTokenizer.from_pretrained("SemViQA/tc-infoxlm-viwikifc") model = ClaimModelForClassification.from_pretrained("SemViQA/tc-infoxlm-viwikifc") claim = "Chiến tranh với Campuchia đã kết thúc trước khi Việt Nam thống nhất." evidence = "Sau khi thống nhất, Việt Nam tiếp tục gặp khó khăn do sự sụp đổ và tan rã của đồng minh Liên Xô cùng Khối phía Đông, các lệnh cấm vận của Hoa Kỳ, chiến tranh với Campuchia, biên giới giáp Trung Quốc và hậu quả của chính sách bao cấp sau nhiều năm áp dụng." inputs = tokenizer( claim, evidence, truncation="only_second", add_special_tokens=True, max_length=256, padding='max_length', return_attention_mask=True, return_token_type_ids=False, return_tensors='pt', ) labels = ["NEI", "SUPPORTED", "REFUTED"] with torch.no_grad(): outputs = model(**inputs) logits = outputs["logits"] probabilities = F.softmax(logits, dim=1).squeeze() for i, (label, prob) in enumerate(zip(labels, probabilities.tolist()), start=1): print(f"{i}) {label} {prob:.4f}") # 1) NEI 0.0001 # 2) SUPPORTED 0.0001 # 3) REFUTED 0.9998 ``` ## **Evaluation Results** SemViQA-TC is one of the key components of the two-step classification (TVC) approach in the SemViQA system. SemViQA-TC achieved impressive results on the test set, demonstrating accurate and efficient classification capabilities. The detailed evaluation of SemViQA-TC is presented in the table below.
Method ViWikiFC
ER VC Strict Acc VC Acc ER Acc Time (s)
TF-IDF InfoXLMlarge 75.56 82.21 90.15 131
XLM-Rlarge 76.47 82.78 90.15 134
Ernie-Mlarge 75.56 81.83 90.15 144
BM25 InfoXLMlarge 70.44 79.01 83.50 130
XLM-Rlarge 70.97 78.91 83.50 132
Ernie-Mlarge 70.21 78.29 83.50 141
SBert InfoXLMlarge 74.99 81.59 89.72 195
XLM-Rlarge 75.80 82.35 89.72 194
Ernie-Mlarge 75.13 81.44 89.72 203
QA-based approaches VC
ViMRClarge InfoXLMlarge 77.28 81.97 92.49 3778
XLM-Rlarge 78.29 82.83 92.49 3824
Ernie-Mlarge 77.38 81.92 92.49 3785
InfoXLMlarge InfoXLMlarge 78.14 82.07 93.45 4092
XLM-Rlarge 79.20 83.07 93.45 4096
Ernie-Mlarge 78.24 82.21 93.45 4102
LLM
Qwen2.5-1.5B-Instruct 51.03 65.18 78.96 7665
Qwen2.5-3B-Instruct 44.38 62.31 71.35 12123
LLM VC
Qwen2.5-1.5B-Instruct InfoXLMlarge 66.14 76.47 78.96 7788
XLM-Rlarge 67.67 78.10 78.96 7789
Ernie-Mlarge 66.52 76.52 78.96 7794
Qwen2.5-3B-Instruct InfoXLMlarge 59.88 72.50 71.35 12246
XLM-Rlarge 60.74 73.08 71.35 12246
Ernie-Mlarge 60.02 72.21 71.35 12251
SER Faster (ours) TVC (ours)
TF-IDF + ViMRClarge Ernie-Mlarge 79.44 82.93 94.60 410
TF-IDF + InfoXLMlarge Ernie-Mlarge 79.77 83.07 95.03 487
SER (ours) TVC (ours)
TF-IDF + ViMRClarge InfoXLMlarge 80.25 83.84 94.69 2731
XLM-Rlarge 80.34 83.64 94.69 2733
Ernie-Mlarge 79.53 82.97 94.69 2733
TF-IDF + InfoXLMlarge InfoXLMlarge 80.68 83.98 95.31 3860
XLM-Rlarge 80.82 83.88 95.31 3843
Ernie-Mlarge 80.06 83.17 95.31 3891
## **Citation** If you use **SemViQA-TC** in your research, please cite: ```bibtex @misc{tran2025semviqasemanticquestionanswering, title={SemViQA: A Semantic Question Answering System for Vietnamese Information Fact-Checking}, author={Dien X. Tran and Nam V. Nguyen and Thanh T. Tran and Anh T. Hoang and Tai V. Duong and Di T. Le and Phuc-Lu Le}, year={2025}, eprint={2503.00955}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2503.00955}, } ``` 🔗 **Paper Link:** [SemViQA on arXiv](https://arxiv.org/abs/2503.00955) 🔗 **Source Code:** [GitHub - SemViQA](https://github.com/DAVID-NGUYEN-S16/SemViQA) ## About *Built by Dien X. Tran* [![LinkedIn](https://img.shields.io/badge/LinkedIn-Profile-blue?logo=linkedin)](https://www.linkedin.com/in/xndien2004/) For more details, visit the project repository. [![GitHub stars](https://img.shields.io/github/stars/DAVID-NGUYEN-S16/SemViQA?style=social)](https://github.com/DAVID-NGUYEN-S16/SemViQA)