Description
We scraped out articles with comments from news articles of top 10 publishers for both countries of analysis, and the purpose was to see which elements in the articles attract readers to leave comment or to carry out discussion about nuclear reactor. Elements are divided into place / figure / organization, every flow of visualization represents the number of articles mentioning the element.
It can be found that in the figure Chinese readers are more interested in scientific researches, while American readers are mostly interested in major national politicians such as Trump, Shinzo Abe, Hillary Clinton; in terms of place, Chinese readers focus on Japan region most, followed by France and former Soviet Union region; US readers focus on widely such as Syria ,DPRK , Korea ,Russia , India ,China ,Japan ,European countries; in terms of organization, China and US readers are commonly interested in TEPCO,Toshiba,Westhouse and official nuclear energy organization, and IS terrorist organization also interests US readers in nuclear issue.
Protocol
For collected articles with comments of China and US in the two periods, we directly put the links of US news into Entity Extraction Too of Aylian to extract place / person / organization elements, the text of Chinese news was translated into English with Google translate first, then put their English text into Aylian’s tool to extract element.
After the result was generated, we collected the mark and counting of every element to excel, scraped out top 10 entities from place/ person/ organization respectively, using rawgraphic.io to visualize and putting them into illustrator for adjustment.
Data
Data source: Google News
, Baidu News
Aylien
Download data(All comments), Download data (Classify)
Aylien doesn’t work for chinese links so we had to use google translate to translate the content into English language.