Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Fine-grained Narrative Classification in Biased News Articles

Zeba Afroz, Harsh Vardhan, Pawan Bhakuni, Aanchal Punia, Rajdeep Kumar, Md. Shad Akhtar

TL;DR

This work tackles propaganda analysis in biased news by introducing a fine-grained narrative classification task and the ideologically grounded INDI-PROP dataset, covering bias, event-specific narratives, and persuasive techniques across India’s CAA and Farmers' protest. It proposes two multi-hop prompting frameworks, FANTA for bias and narrative classification and TPTC for persuasive technique identification, and demonstrates that hierarchical, context-aware reasoning improves performance over traditional fine-tuned models. The dataset and models enable narrative-aware propaganda analysis in a non-Western, multilingual context and provide benchmarks for bias, framing, and persuasion research. Overall, the study shows that decomposing narrative and persuasion into hierarchical steps helps LLMs bridge micro-level cues with macro-level ideological narratives, advancing computational social science of media framing.

Abstract

Narratives are the cognitive and emotional scaffolds of propaganda. They organize isolated persuasive techniques into coherent stories that justify actions, attribute blame, and evoke identification with ideological camps. In this paper, we propose a novel fine-grained narrative classification in biased news articles. We also explore article-bias classification as the precursor task to narrative classification and fine-grained persuasive technique identification. We develop INDI-PROP, the first ideologically grounded fine-grained narrative dataset with multi-level annotation for analyzing propaganda in Indian news media. Our dataset INDI-PROP comprises 1,266 articles focusing on two polarizing socio-political events in recent times: CAA and the Farmers' protest. Each article is annotated at three hierarchical levels: (i) ideological article-bias (pro-government, pro-opposition, neutral), (ii) event-specific fine-grained narrative frames anchored in ideological polarity and communicative intent, and (iii) persuasive techniques. We propose FANTA and TPTC, two GPT-4o-mini guided multi-hop prompt-based reasoning frameworks for the bias, narrative, and persuasive technique classification. FANTA leverages multi-layered communicative phenomena by integrating information extraction and contextual framing for hierarchical reasoning. On the other hand, TPTC adopts systematic decomposition of persuasive cues via a two-stage approach. Our evaluation suggests substantial improvement over underlying baselines in each case.

Fine-grained Narrative Classification in Biased News Articles

TL;DR

This work tackles propaganda analysis in biased news by introducing a fine-grained narrative classification task and the ideologically grounded INDI-PROP dataset, covering bias, event-specific narratives, and persuasive techniques across India’s CAA and Farmers' protest. It proposes two multi-hop prompting frameworks, FANTA for bias and narrative classification and TPTC for persuasive technique identification, and demonstrates that hierarchical, context-aware reasoning improves performance over traditional fine-tuned models. The dataset and models enable narrative-aware propaganda analysis in a non-Western, multilingual context and provide benchmarks for bias, framing, and persuasion research. Overall, the study shows that decomposing narrative and persuasion into hierarchical steps helps LLMs bridge micro-level cues with macro-level ideological narratives, advancing computational social science of media framing.

Abstract

Narratives are the cognitive and emotional scaffolds of propaganda. They organize isolated persuasive techniques into coherent stories that justify actions, attribute blame, and evoke identification with ideological camps. In this paper, we propose a novel fine-grained narrative classification in biased news articles. We also explore article-bias classification as the precursor task to narrative classification and fine-grained persuasive technique identification. We develop INDI-PROP, the first ideologically grounded fine-grained narrative dataset with multi-level annotation for analyzing propaganda in Indian news media. Our dataset INDI-PROP comprises 1,266 articles focusing on two polarizing socio-political events in recent times: CAA and the Farmers' protest. Each article is annotated at three hierarchical levels: (i) ideological article-bias (pro-government, pro-opposition, neutral), (ii) event-specific fine-grained narrative frames anchored in ideological polarity and communicative intent, and (iii) persuasive techniques. We propose FANTA and TPTC, two GPT-4o-mini guided multi-hop prompt-based reasoning frameworks for the bias, narrative, and persuasive technique classification. FANTA leverages multi-layered communicative phenomena by integrating information extraction and contextual framing for hierarchical reasoning. On the other hand, TPTC adopts systematic decomposition of persuasive cues via a two-stage approach. Our evaluation suggests substantial improvement over underlying baselines in each case.

Paper Structure

This paper contains 36 sections, 4 figures, 8 tables.

Table of Contents

  1. Introduction
  2. Contributions:
  3. Related Work
  4. Novelty in our work:
  5. Dataset
  6. Data Collection
  7. Data Cleaning:
  8. Annotation Guidelines
  9. Article-Bias Guidelines:
  10. Fine-Grain Narrative Guidelines:
  11. Pro-Govt Aligned Narratives (Farmer’s protest):
  12. Pro-Opp Aligned Narratives (Farmer’s protest):
  13. Pro-Govt Aligned Narratives (CAA):
  14. Pro-Opp Aligned Narratives (CAA):
  15. Persuasive Technique Guidelines:
  16. ...and 21 more sections

Figures (4)

  • Figure 1: Multi-level schema of INDI-PROP.
  • Figure 2: Narrative Taxonomy for two socio-political movements: Farmers' Protest and CAA/NRC. Taxonomy is organized along the ideological orientation of Pro-Govt and Pro-Opp.
  • Figure 3: Overview of the proposed models architecture: The upper pipeline illustrates FANTA for the Bias and Fine-Grain Narrative Classification process. The lower pipeline illustrates TPTC for the Persuasive Technique Classification process.
  • Figure 4: Confusion matrix for Article-Bias Classification.