Data visualization of texts in the transitions of framing Indochinese refugees by Japanese television documentaries

In contrast to the previous studies that have conducted visual analyses of refugee documentaries, the aim of this study is to provide new insight into how the refugee crisis in Indochina has been portrayed in the Japanese media through documentaries by text mining narratives of the television program scripts. Out of the total of all 35 television documentaries that have been broadcasted by NHK in the past, this study sampled ten television documentaries, the oldest going back to the 1970s. In order to historically view how the portrayal of Indochinese refugees have changed over the years and how this might affect how Japanese perceive refugees, the following were sampled for this paper: one from 1977, four from the 1980s, two from the 1990s, two from the 2000s and one from 2014. The narration scripts of these documentary programs were re-created by hand into data, onsite at the NHK Archives in Tokyo. The data analysis was done through text mining by R. From the text-mining analysis and data visualization by weighted tag clouds and co-occurrence network maps of all ten documentaries, certain characteristics of the dominant refugee media agenda of each decade were clearly observed. One noteworthy finding is the general trend of how initially, refugees were seen as an alien collective, and gradually over the decades, the media made as transition of showing them more as individuals, families and finally as a part of the Japanese community.