Hybrid Annotation for Propaganda Detection: Integrating LLM Pre-Annotations with Human Intelligence

TL;DR

This paper presents a hybrid annotation framework combining human expertise and Large Language Model assistance to improve propaganda detection, achieving higher consistency and efficiency in labeling social media data.

Contribution

It introduces a hierarchical taxonomy, an LLM-assisted pre-annotation pipeline, and a knowledge distillation approach to enhance scalable propaganda annotation.

Findings

Improved inter-annotator agreement with LLM assistance

Significant reduction in annotation time

Effective training of smaller models on high-quality LLM-generated data

Abstract

Propaganda detection on social media remains challenging due to task complexity and limited high-quality labeled data. This paper introduces a novel framework that combines human expertise with Large Language Model (LLM) assistance to improve both annotation consistency and scalability. We propose a hierarchical taxonomy that organizes 14 fine-grained propaganda techniques into three broader categories, conduct a human annotation study on the HQP dataset that reveals low inter-annotator agreement for fine-grained labels, and implement an LLM-assisted pre-annotation pipeline that extracts propagandistic spans, generates concise explanations, and assigns local labels as well as a global label. A secondary human verification study shows significant improvements in both agreement and time-efficiency. Building on this, we fine-tune smaller language models (SLMs) to perform structured annotation. Instead of fine-tuning on human annotations, we train on high-quality LLM-generated data, allowing a large model to produce these annotations and a smaller model to learn to generate them via knowledge distillation. Our work contributes towards the development of scalable and robust propaganda detection systems, supporting the idea of transparent and accountable media ecosystems in line with SDG 16. The code is publicly available at our GitHub repository.