Synthetically generated text for supervised text analysis

TL;DR

This paper explores the use of large language models to generate synthetic text for supervised political text analysis, addressing data scarcity, privacy, and copyright issues.

Contribution

It introduces a framework for controlled synthetic text generation, discusses ethical considerations, and demonstrates practical applications in political science research.

Findings

Synthetic tweets effectively describe real-world events.

Synthetic data improves training for event detection systems.

Multilingual synthetic corpora aid in classifying political language.

Abstract

Supervised text models are a valuable tool for political scientists but present several obstacles to their use, including the expense of hand-labeling documents, the difficulty of retrieving rare relevant documents for annotation, and copyright and privacy concerns involved in sharing annotated documents. This article proposes a partial solution to these three issues, in the form of controlled generation of synthetic text with large language models. I provide a conceptual overview of text generation, guidance on when researchers should prefer different techniques for generating synthetic text, a discussion of ethics, and a simple technique for improving the quality of synthetic text. I demonstrate the usefulness of synthetic text with three applications: generating synthetic tweets describing the fighting in Ukraine, synthetic news articles describing specified political events for training an event detection system, and a multilingual corpus of populist manifesto statements for training a sentence-level populism classifier.