Semantic structure of disputed issues in Japanese and South Korean newspapers: A text mining approach using Word2Vec

Tetsuro Kobayashi, Yuki Ogawa

Research output: Conference PapersRGC 32 - Refereed conference paper (without host publication)peer-review

Abstract

International disputes are difficult to resolve, partly because the disputed issues are framed differently by the news media of the countries involved. Therefore, it is essential to compare the news coverage of countries having disagreements by using the identical methodology and investigate differences in the framing of such conflicts. To investigate territorial and historical disputes between Japan and Korea, we quantitatively analyzed newspaper articles published between 2011 and 2016 in liberal and conservative newspapers of the two countries (Asahi, Yomiuri, Hankyoreh, and Chosun). More specifically, we used Word2Vec, which expresses the meaning of words as vectors, and investigated the semantic structure of issues related to Takeshima/Dokdo and comfort women in news articles. Results indicated that framing of these issues differed significantly between the Japanese and the South Korean press, as well as between the liberal and the conservative media in each country. We suggest that these differences could potentially make reconciliation between Japan and Korea more challenging.
Original languageEnglish
Publication statusPublished - Jan 2019
Event6th Asian Political Methodology Meeting (Asia PolMeth2019) - Doshisha University, Kyoto, Japan
Duration: 5 Jan 20196 Jan 2019
https://polmeth.org/blog/2019-asian-political-methodology-meeting
https://polmeth.org/asia

Conference

Conference6th Asian Political Methodology Meeting (Asia PolMeth2019)
Abbreviated titleAsia PolMeth 2019
Country/TerritoryJapan
CityKyoto
Period5/01/196/01/19
Internet address

Bibliographical note

Information for this record is supplemented by the author(s) concerned.

Fingerprint

Dive into the research topics of 'Semantic structure of disputed issues in Japanese and South Korean newspapers: A text mining approach using Word2Vec'. Together they form a unique fingerprint.

Cite this