DariMis: Harm-Aware Modeling for Dari Misinformation Detection on YouTube

• URL: http://arxiv.org/abs/2603.22977v1
• Date: Tue, 24 Mar 2026 09:13:40 GMT
• Title: DariMis: Harm-Aware Modeling for Dari Misinformation Detection on YouTube
• Abstract summary: Dari, the primary language of Afghanistan, is spoken by tens of millions of people yet remains largely absent from the misinformation detection literature.<n>We address this gap with DariMis, the first manually annotated dataset of 9,224 Dari-language YouTube videos, labeled across two dimensions: Information Type (Misinformation, Partly True, True) and Harm Level (Low, Medium, High)<n>We propose a pair-input encoding strategy that represents the video title and description as separate BERT segment inputs, explicitly modeling the semantic relationship between headline claims and body content.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Dari, the primary language of Afghanistan, is spoken by tens of millions of people yet remains largely absent from the misinformation detection literature. We address this gap with DariMis, the first manually annotated dataset of 9,224 Dari-language YouTube videos, labeled across two dimensions: Information Type (Misinformation, Partly True, True) and Harm Level (Low, Medium, High). A central empirical finding is that these dimensions are structurally coupled, not independent: 55.9 percent of Misinformation carries at least Medium harm potential, compared with only 1.0 percent of True content. This enables Information Type classifiers to function as implicit harm-triage filters in content moderation pipelines. We further propose a pair-input encoding strategy that represents the video title and description as separate BERT segment inputs, explicitly modeling the semantic relationship between headline claims and body content, a key signal of misleading information. An ablation study against single-field concatenation shows that pair-input encoding yields a 7.0 percentage point gain in Misinformation recall (60.1 percent to 67.1 percent), the safety-critical minority class, despite modest overall macro F1 differences (0.09 percentage points). We benchmark a Dari/Farsi-specialized model (ParsBERT) against XLM-RoBERTa-base; ParsBERT achieves the best test performance with accuracy of 76.60 percent and macro F1 of 72.77 percent. Bootstrap 95 percent confidence intervals are reported for all metrics, and we discuss both the practical significance and statistical limitations of the results.

Related papers

• SAFIRE: Safety-Critical Benchmark for Fine-grained Fire and Smoke Understanding in Multimodal LLMs [28.105080919216167]
  Fire-smoke understanding is central to public safety and disaster response.<n>Most existing benchmarks lack diverse real-world scenarios and context-aware evaluation.<n>We introduce SAFIRE, a large-scale benchmark for fire-smoke understanding in MLLMs.
  arXiv  Detail & Related papers  (2026-09-07T17:59:50Z)
• FedPref: Federated Preference Learning for Structured Radiology Report Extraction [69.11739400975445]
  Radiology reports describe findings and locations in free text, but downstream search and analysis require these relations in a fixed schema.<n>We introduce FedPref: frozen public language models propose alternative extractions, local annotations rank them, and sites collaboratively train pooled compact Qwen3-8B adapters.<n>On development data from six simulated hospitals with unequal data volume and disease prevalence, FedPref improves client-mean F1 by 2.49 points and worst-site F1 by 9.10 points compared with training each site in isolation.
  arXiv  Detail & Related papers  (2026-08-17T12:29:42Z)
• DIA-HARM: Dialectal Disparities in Harmful Content Detection Across 50 English Dialects [6.107850985025956]
  We present DIA-HARM, the first benchmark for evaluating disinformation detection across 50 English dialects.<n>Human-written dialectal content degrades detection by 1.4-3.6% F1, while AI-generated content remains stable.<n>We release the DIA-HARM framework, D3 corpus, and evaluation tools.
  arXiv  Detail & Related papers  (2026-04-07T01:43:48Z)
• Membership Inference Attacks against Large Audio Language Models [50.84901010528239]
  We present the first systematic Membership Inference Attack (MIA) evaluation of Large Audio Language Models (LALMs)<n>Using a multi-modal blind baseline based on textual, spectral, and prosodic features, we demonstrate that common speech datasets exhibit near-perfect train/test separability.<n>These findings establish a principled standard for auditing LALMs beyond spurious correlations.
  arXiv  Detail & Related papers  (2026-03-30T12:45:28Z)
• Bielik Guard: Efficient Polish Language Safety Classifiers for LLM Content Moderation [0.4349640169711269]
  Bielik Guard is a family of compact Polish language safety classifiers.<n>Fine-tuned on a community-annotated dataset of 6,885 Polish texts, these models classify content across five safety categories: Hate/Aggression, Vulgarities, Sexual Content, Crime, and Self-Harm.
  arXiv  Detail & Related papers  (2026-02-08T12:57:04Z)
• GerAV: Towards New Heights in German Authorship Verification using Fine-Tuned LLMs on a New Benchmark [20.533795195003286]
  Authorship verification (AV) is the task of determining whether two texts were written by the same author.<n>GerAV is a comprehensive benchmark for German AV comprising over 600k labeled text pairs.
  arXiv  Detail & Related papers  (2026-01-20T08:08:18Z)
• HyFormer-Net: A Synergistic CNN-Transformer with Interpretable Multi-Scale Fusion for Breast Lesion Segmentation and Classification in Ultrasound Images [0.0]
  HyFormer-Net is a hybrid CNN-Transformer for simultaneous segmentation and classification with intrinsic interpretability.<n>On the BUSI dataset, HyFormer-Net achieves Dice Score 0.761 +/- 0.072 and accuracy 93.2%, outperforming U-Net, Attention U-Net, and TransUNet.<n>We conduct the first cross-dataset generalization study for hybrid CNN-Transformers in breast ultrasound.
  arXiv  Detail & Related papers  (2025-11-02T17:06:12Z)
• Benchmarking Reasoning Robustness in Large Language Models [76.79744000300363]
  We find significant performance degradation on novel or incomplete data.<n>These findings highlight the reliance on recall over rigorous logical inference.<n>This paper introduces a novel benchmark, termed as Math-RoB, that exploits hallucinations triggered by missing information to expose reasoning gaps.
  arXiv  Detail & Related papers  (2025-03-06T15:36:06Z)
• RideKE: Leveraging Low-Resource, User-Generated Twitter Content for Sentiment and Emotion Detection in Kenyan Code-Switched Dataset [4.142287865325998]
  We analyze Kenyan code-switched data and evaluate four state-of-the-art (SOTA) transformer-based pretrained models for sentiment and emotion classification.<n>For sentiment analysis, XLM-R supervised model achieves the highest accuracy (69.2%) and F1 score (66.1%), XLM-R semi-supervised (67.2% accuracy, 64.1% F1 score)<n>In emotion analysis, DistilBERT supervised leads in accuracy (59.8%) and F1 score (31%), mBERT semi-supervised (accuracy (59% and F1 score 26.5
  arXiv  Detail & Related papers  (2025-02-10T06:18:07Z)
• Exploring Response Uncertainty in MLLMs: An Empirical Evaluation under Misleading Scenarios [49.53589774730807]
  Multimodal large language models (MLLMs) have recently achieved state-of-the-art performance on tasks ranging from visual question answering to video understanding.<n>We reveal a response uncertainty phenomenon: twelve state-of-the-art open-source MLLMs overturn a previously correct answer in 65% of cases after receiving a single deceptive cue.
  arXiv  Detail & Related papers  (2024-11-05T01:11:28Z)
• When Does Monolingual Data Help Multilingual Translation: The Role of Domain and Model Scale [73.69252847606212]
  We examine how denoising autoencoding (DAE) and backtranslation (BT) impact machine translation (MMT) We find that monolingual data generally helps MMT, but models are surprisingly brittle to domain mismatches, especially at smaller model scales. As scale increases, DAE transitions from underperforming the parallel-only baseline at 90M to converging with BT performance at 1.6B, and even surpassing it in low-resource.
  arXiv  Detail & Related papers  (2023-05-23T14:48:42Z)
• UATVR: Uncertainty-Adaptive Text-Video Retrieval [90.8952122146241]
  A common practice is to transfer text-video pairs to the same embedding space and craft cross-modal interactions with certain entities. We propose an Uncertainty-language Text-Video Retrieval approach, termed UATVR, which models each look-up as a distribution matching procedure.
  arXiv  Detail & Related papers  (2023-01-16T08:43:17Z)
• HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units [81.53783563025084]
  We propose an offline clustering step to provide aligned target labels for a BERT-like prediction loss. A key ingredient of our approach is applying the prediction loss over the masked regions only. HuBERT shows up to 19% and 13% relative WER reduction on the more challenging dev-other and test-other evaluation subsets.
  arXiv  Detail & Related papers  (2021-06-14T14:14:28Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.