# HiFACTMix: A Code-Mixed Benchmark and Graph-Aware Model for EvidenceBased Political Claim Verification in Hinglish

arXiv: https://arxiv.org/abs/2508.10001 · HTML: https://mlsanity.com/paper/2508.10001/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 6 of its 7 references that are in the map.
- First posted on arXiv 4 Aug 2025.

## Builds on (most cited first, 6 of 6)

- [Molecular Facts: Desiderata for Decontextualization in LLM Fact Verification](https://mlsanity.com/paper/2406.20079/) (2024)
- [Towards Effective Extraction and Evaluation of Factual Claims](https://mlsanity.com/paper/2502.10855/) (2025)
- [FACT-AUDIT: An Adaptive Multi-Agent Framework for Dynamic Fact-Checking Evaluation of Large Language Models](https://mlsanity.com/paper/2502.17924/) (2025)
- [VeriFact: Enhancing Long-Form Factuality Evaluation with Refined Fact Extraction and Reference Facts](https://mlsanity.com/paper/2505.09701/) (2025)
- [Beyond Translation: LLM-Based Data Generation for Multilingual Fact-Checking](https://mlsanity.com/paper/2502.15419/) (2025)
- [FACTS&EVIDENCE: An Interactive Tool for Transparent Fine-Grained Factual Verification of Machine-Generated Text](https://mlsanity.com/paper/2503.14797/) (2025)

## Abstract (arXiv)

Fact-checking in code-mixed, low-resource languages such as Hinglish remains an underexplored challenge in natural language processing. Existing fact-verification systems largely focus on high-resource, monolingual settings and fail to generalize to real-world political discourse in linguistically diverse regions like India. Given the widespread use of Hinglish by public figures, particularly political figures, and the growing influence of social media on public opinion, there's a critical need for robust, multilingual and context-aware fact-checking tools. To address this gap a novel benchmark HiFACT dataset is introduced with 1,500 realworld factual claims made by 28 Indian state Chief Ministers in Hinglish, under a highly code-mixed low-resource setting. Each claim is annotated with textual evidence and veracity labels. To evaluate this benchmark, a novel graphaware, retrieval-augmented fact-checking model is proposed that combines multilingual contextual encoding, claim-evidence semantic alignment, evidence graph construction, graph neural reasoning, and natural language explanation generation. Experimental results show that HiFACTMix outperformed accuracy in comparison to state of art multilingual baselines models and provides faithful justifications for its verdicts. This work opens a new direction for multilingual, code-mixed, and politically grounded fact verification research.