UKElectionNarratives: A Dataset of Misleading Narratives Surrounding Recent UK General Elections
🤖 AI Summary
This study investigates the mechanisms through which misleading narratives influence public opinion during UK general elections. Addressing the lack of systematic taxonomies and empirical data on electoral disinformation in Europe, the work introduces the first taxonomy of misleading narratives specifically designed for European elections. Leveraging this taxonomy, the authors construct UKElectionNarratives—the first temporally aligned, manually annotated dataset of multi-source misleading narratives covering the 2019 and 2024 UK general elections. They release a detailed coding manual and a standardized benchmark evaluation protocol. Detection experiments using state-of-the-art LLMs—including GPT-4o—yield an F1 score of only 68.3%, substantially below human expert performance (89.1%), exposing fundamental limitations of current LLMs in fine-grained political narrative identification. This work establishes a theoretical framework, provides high-quality resources, and sets a methodological benchmark for research on electoral misinformation.
Application Category
📝 Abstract
Misleading narratives play a crucial role in shaping public opinion during elections, as they can influence how voters perceive candidates and political parties. This entails the need to detect these narratives accurately. To address this, we introduce the first taxonomy of common misleading narratives that circulated during recent elections in Europe. Based on this taxonomy, we construct and analyse UKElectionNarratives: the first dataset of human-annotated misleading narratives which circulated during the UK General Elections in 2019 and 2024. We also benchmark Pre-trained and Large Language Models (focusing on GPT-4o), studying their effectiveness in detecting election-related misleading narratives. Finally, we discuss potential use cases and make recommendations for future research directions using the proposed codebook and dataset.
Problem
Research questions and friction points this paper is trying to address.
Detecting misleading narratives in UK elections
Creating a taxonomy for election misinformation
Benchmarking LLMs for narrative detection
Innovation
Methods, ideas, or system contributions that make the work stand out.
First taxonomy of common misleading election narratives
Human-annotated dataset of UK election misleading narratives
Benchmarked LLMs for detecting election misleading narratives
🔎 Similar Papers
No similar papers found.
