Prompt-based contrastive learning to combat the COVID-19 infodemic

The COVID-19 pandemic has brought about an influx of misinformation and disinformation online, especially on social media. The World Health Organization has identified combating this infodemic as one of its top priorities, as false and misleading information can lead to negative consequences, such as the spread of conspiracy theories, false remedies, and xenophobia. This study presents a prompt-based contrastive learning approach that can be employed to address this issue. This method was designed to overcome challenges such as data scarcity and class imbalance commonly found in social media. Fighting the infodemic is modeled as a series of text classification problems in which questions relevant to credibility of the texts, their potential harm to society and the necessity of government intervention need to be answered. Experiments show that prompt-based contrastive learning is effective in assessing the accuracy of COVID-19-related online text. Copyright © 2025 The Author(s), under exclusive licence to Springer Science+Business Media LLC, part of Springer Nature.