Topic Modeling Analysis of the Olympic Discourse in U.S. Mainstream Media 
Yu Dong

School of English Studies Dalian University of Foreign Languages, Liaoning, 116044;

Abstract: Using the Associated Press (AP) as an example, this study aims to uncover the implicit topics, the diachronic distributional features of the topic model,  and the hidden stances and attitudes in Olympic coverage in U.S. mainstream media. Combining the methods of topic modeling and discourse historical approach, the study trained and analyzed topic models based on a corpus of AP Olympic coverage from 2012 to 2022, and produced some useful findings. First, according to the diachronic distributional structure of the model, topics are clustered into two major categories—general and specific topics. Second, the topic model reflected significant attitudinal differences among different targets of the news report. The results of the study may provide a promising direction for combining qualitative and quantitative perspectives in the analysis of media discourse.

Keywords: Olympic coverage; Associated Press, corpus; topic modeling; discourse-historical approach

References

[1]Guo Q, Zhou Y, Jia Z. A framing study of domestic newspaper coverage of the Beijing Olympics [J]. Contemporary Communications, 2008, 3 42-44.

[2]Liu Z. A comparison of the 2008 Beijing Olympic Games coverage in The New York Times and People’s Daily: A framing theory perspective [J]. Journal of Capital University of Physical Education and Sports, 2011, 23(04) 313-316.

[3]Luan Y. The reconstruction of the Olympic communication framework from the perspective of the Rio Olympics [J]. News and Writing, 2016, 9 22-25.

[4]Dong X, Peng S, Wang Q, et al. Agenda setting of English-language media and its implications for national image management: A tracking Study of report themes on five Olympic host cities [J]. China Soft Science, 2008, 6 16-27.

[5]Chen L. The representation and development of China’s image in BBC Olympic coverage: Taking Beijing, London and Rio Olympics as examples [J]. Journal of Beijing Sport University, 2016, 39(12) 37-43.

[6]Zhang T. Consideration and evolution in the narrative words of Olympic Games report in China [J]. Journal of Shandong Sport University, 2018, 34(03) 58-64.

[7]Fairclough N, Wodak R. Critical discourse analysis [M]. Discourse as Social Interaction, 1997, 258-284.

[8]Wodak R. The discourse of politics in action: Politics as usual [M]. Palgrave, 2009. DOI:10.1057/9780230281851.

[9]Reisigl M, Wodak R. The discourse-historical approach (DHA) [M]. Methods of Critical Discourse Studies, 2016, 23-61.

[10]Lee C. How are ‘immigrant workers’ represented in Korean news reporting? A text mining approach to critical discourse analysis [J]. Digital Scholarship in the Humanities, 2019, 34(1) 82-99. DOI:10.1093/LLC/FQY014.

[11]Blei D, Jordan M I. Latent Dirichlet allocation [J]. Journal of Machine Learning Research, 2003, 3(1) 993-1022.

[12]McCallum K. Mallet: Machine learning for language toolkit [EB/OL]. Mallet, 2002.

[13]Mu D. Political intervention in Olympic Games: A case study of US media coverage of the Moscow Summer Olympics and Sochi Winter Olympics [J]. Jianghuai Tribune, 2016, 2 189-192.