# Fine-tuning with Hierarchical Prompting for Robust Propaganda Classification Across Annotation Schemas

arXiv: https://arxiv.org/abs/2605.13663 · HTML: https://mlsanity.com/paper/2605.13663/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 8 of its 31 references that are in the map.
- First posted on arXiv 13 May 2026.

## Builds on (most cited first, 8 of 8)

- [Fine-Grained Analysis of Propaganda in News Articles](https://mlsanity.com/paper/1910.02517/) (2019)
- [SemEval-2020 Task 11: Detection of Propaganda Techniques in News Articles](https://mlsanity.com/paper/2009.02696/) (2020)
- [Can GPT-4 Identify Propaganda? Annotation and Detection of Propaganda Spans in News Articles](https://mlsanity.com/paper/2402.17478/) (2024)
- [Large Language Models for Propaganda Span Annotation](https://mlsanity.com/paper/2311.09812/) (2023)
- [PropaInsight: Toward Deeper Understanding of Propaganda in Terms of Techniques, Appeals, and Intent](https://mlsanity.com/paper/2409.18997/) (2024)
- [PropXplain: Can LLMs Enable Explainable Propaganda Detection?](https://mlsanity.com/paper/2502.16550/) (2025)
- [Check News in One Click: NLP-Empowered Pro-Kremlin Propaganda Detection](https://mlsanity.com/paper/2401.15717/) (2024)
- [Are Large Language Models Good at Detecting Propaganda?](https://mlsanity.com/paper/2505.13706/) (2025)

## Abstract (arXiv)

Propaganda detection in social media is challenging due to noisy, short texts and low annotation agreements. We introduce a new intent-focused taxonomy of propaganda techniques and compare it against an established, higher-agreement schema. Along three dimensions (model portfolio, schema effects, and prompting strategy) we evaluate the taxonomies as a classification task with the help of four language models (GPT-4.1-nano, Phi-4 14B, Qwen2.5-14B, Qwen3-14B). Our results show that fine-tuning is essential, since it transforms weak zero-shot baselines into competitive systems and reveals methodological differences that are hidden using base models. Across schemas, the Qwen models achieve the strongest overall performance, and Phi-4 14B consistently outperforms GPT-4.1-nano. Our hierarchical prompting method (HiPP), which predicts fine-grained techniques before aggregating them, is especially beneficial after fine-tuning and on the more ambiguous, low-agreement taxonomy, while remaining competitive on the simpler schema. The HQP dataset, annotated with the new intent-based labels, provides a richer lens on propaganda's strategic goals and a challenging benchmark for future work on robust, real-world detection.