Network-informed Prompt Engineering against Organized Astroturf Campaigns under Extreme Class Imbalance

TL;DR

The paper tackles the problem of detecting coordinated astroturf campaigns on social media under extreme class imbalance. It introduces a training-free framework that encodes both textual content and the propagation graph into prompts for frozen LLMs, augmented by Balanced Retrieval-Augmented Generation to mitigate data scarcity. On a FakeNewsNet-based dataset from the 2016 US election, the approach outperforms traditional graph-based baselines, achieving notable gains in recall, precision, F1, and ROC AUC, with Balanced RAG and few-shot prompting yielding the best results. By avoiding model fine-tuning and exploiting graph-to-text encodings coupled with targeted retrieval, the method offers a scalable, robust solution for real-world disinformation detection under imbalanced conditions.

Abstract

Detecting organized political campaigns is of paramount importance in fighting against disinformation on social media. Existing approaches for the identification of such organized actions employ techniques mostly from network science, graph machine learning and natural language processing. Their ultimate goal is to analyze the relationships and interactions (e.g. re-posting) among users and the textual similarities of their posts. Despite their effectiveness in recognizing astroturf campaigns, these methods face significant challenges, notably the class imbalance in available training datasets. To mitigate this issue, recent methods usually resort to data augmentation or increasing the number of positive samples, which may not always be feasible or sufficient in real-world settings. Following a different path, in this paper, we propose a novel framework for identifying astroturf campaigns based solely on large language models (LLMs), introducing a Balanced Retrieval-Augmented Generation (Balanced RAG) component. Our approach first gives both textual information concerning the posts (in our case tweets) and the user interactions of the social network as input to a language model. Then, through prompt engineering and the proposed Balanced RAG method, it effectively detects coordinated disinformation campaigns on X (Twitter). The proposed framework does not require any training or fine-tuning of the language model. Instead, by strategically harnessing the strengths of prompt engineering and Balanced RAG, it facilitates LLMs to overcome the effects of class imbalance and effectively identify coordinated political campaigns. The experimental results demonstrate that by incorporating the proposed prompt engineering and Balanced RAG methods, our framework outperforms the traditional graph-based baselines, achieving 2x-3x improvements in terms of precision, recall and F1 scores.

Figures (1)

• Figure 1: Comparison of different metrics (precision, recall, F1 score, and ROC AUC) across various (Llama) model sizes. The graph illustrates how each metric changes as the number of model parameters increases.

Theorems & Definitions (3)

• definition 1
• definition 2
• definition 3