Media Representation of China in COVID-19 Reports: Text Mining of the Language of New York Times

Textual mining, an application of natural language processing and analytical methods, effectively turns text into data, making machine analytics possible, especially in the field of textual framing and sentiment analysis, traditionally classified manually by researchers which unavoidably involves tremendous manpower and time. This study examines the themes and sentiments of news coverage of China against the backdrop of Covid-19 in the New York Times (NYT) ranging from January 2020 to January 2021 by employing the LDA topic modeling textbfand (Natural Language Toolkit) Vader SentimentAnalysertextbf. The result of a combination of quantitative and qualitative analysis reveals the foci and attitudes of NYT and highlights the selection, emphasis and exclusion practices in this Western media. The study thus broadens the scope of existing content analyses of the image of China and contributes to the exploratory application of text mining techniques in media and linguistic studies.