Abstract
The COVID-19 infodemic, characterized by the rapid spread of misinformation and unverified claims related to the pandemic, presents a significant challenge. This paper presents a comparative analysis of the COVID-19 infodemic in the English and Chinese languages, utilizing textual data extracted from social media platforms. To ensure a balanced representation, two infodemic datasets were created by augmenting previously collected social media textual data. Through word frequency analysis, the thirty-five most frequently occurring infodemic words are identified, shedding light on prevalent discussions surrounding the infodemic. Moreover, topic clustering analysis uncovers thematic structures and provides a deeper understanding of primary topics within each language context. Additionally, sentiment analysis enables comprehension of the emotional tone associated with COVID-19 information on social media platforms in English and Chinese. This research contributes to a better understanding of the COVID-19 infodemic phenomenon and can guide the development of strategies to combat misinformation during public health crises across different languages.
Introduction
The "infodemic"—the overwhelming flood of information, including misinformation, that accompanies a health crisis—has been a defining feature of the COVID-19 pandemic. While much research has focused on the technical detection of fake news, there is a critical need to understand the underlying characteristics of this phenomenon across different linguistic and cultural landscapes. Misinformation is not a universal constant; it is shaped by the cultural norms, political climates, and online behaviors of the communities in which it circulates. This study provides a cross-linguistic analysis of how the infodemic manifested on English and Chinese social media platforms, offering insights into the distinct ways these two major global languages processed the crisis.
Methodology
The researchers utilized Natural Language Processing (NLP) to analyze curated datasets from English and Chinese social media. The methodology involved three primary techniques:
- Word Frequency Analysis: Identifying the top 35 nouns and gerunds in each language to capture core themes.
- Topic Clustering (LDA): Applying Latent Dirichlet Allocation to reveal the thematic structure and diversity of conversations.
- Sentiment Analysis: Categorizing posts into positive, neutral, or negative emotional tones using specialized tools for each language (Monkeylearn for English and ROST_CM6 for Chinese).
The datasets were refined with the assistance of healthcare professionals to ensure accurate classification into "True," "False," and "Uncertain" categories, resulting in 2,490 English records and 1,055 Chinese records.
Word Frequency and Cultural Signatures
The word frequency analysis revealed striking differences in the focus of public discourse. In English-speaking contexts, the infodemic was highly politicized, with "President Donald Trump" appearing as a central figure. In contrast, Chinese discourse was more oriented toward medical expertise and regional impacts, with "Zhong Nanshan," a prominent medical expert, and cities like "Wuhan" being frequently mentioned.
Furthermore, the "Mask" cultural divide was evident; the term "口罩" (mask) was significantly more prominent in Chinese word clouds, reflecting early and widespread adoption in China. English posts often exhibited a "temporal" focus (e.g., "Today," "Yesterday"), indicating a demand for real-time updates, whereas Chinese posts maintained a "regional" focus, highlighting the localized nature of the crisis response in cities like Beijing and Shanghai.
Topic Clustering and Sentiment Trends
The study found that English discussions were more concentrated, with only four major topics identified, while Chinese discussions were more fragmented, spanning thirteen distinct topics. Interestingly, the "False" news category in English contained a greater diversity of topics than its Chinese counterpart, suggesting a wider variety of conspiracy theories and false narratives in English-speaking social media.
Sentiment analysis showed a dominance of negative emotions across both languages, accounting for over 50% of the data. However, English-speaking users appeared more polarized, with a higher proportion of positive sentiment (31.12%) but very little neutral sentiment (8.92%). Chinese records were more balanced between positive (25.69%) and neutral (23.60%) tones. In both languages, "False" news was consistently the most negative, underscoring the role of misinformation in fueling fear and anger.
Conclusion
This comparative analysis highlights that while the COVID-19 infodemic is a global phenomenon, its characteristics are deeply rooted in linguistic and cultural contexts. Understanding these differences—such as the politicization of English discourse versus the medical/regional focus in Chinese—is essential for developing effective, localized strategies to combat misinformation. Future research will focus on the dynamic evolution of these patterns over time and the propagation mechanisms of misinformation across different platforms.
Summary updated