論文

査読有り
2013年3月

Twitterにおけるバーストの生起要因と類型化に関する分析

情報社会学会誌
  • 水沼 友宏
  • ,
  • 池内 淳
  • ,
  • 山本 修平
  • ,
  • 山口 裕太郎
  • ,
  • 佐藤 哲司
  • ,
  • 島田 諭

7
2
開始ページ
41
終了ページ
50
記述言語
日本語
掲載種別
研究論文(学術雑誌)
DOI
10.15099/00000836
出版者・発行元
情報社会学会

Today, there are many different social communication platforms (eg. Facebook,Twitter, LinkedIn Google+,mixi and so forth).Internet users are spending more time on social media. Twitter which is a kind of microbloglaunched in October 2006, hasbecome one of the most popular social communication tool in Japan. Because of Twitter's characteristic features such asimmediacy, easiness and bi-directionality, Twitter's timeline instantly reflects the real world. Once a major event happens,thenumber of tweets rapidly increases. In this article, this phenomenon is defined as 'burst'. The authors gathered Japanese tweetson public timeline from Twitter API over a period of four months starting from March 14, 2012 to July 13, 2012. We collected atotal of 1,379,844,343 posts from 58,866,923 distinct users. This study shows that tweets in bursty term are more retweet, lessreply and less charcters than usual. We have detected four types of bursts by cluster analysis. And the Japanese scale ofearthquake and the distance from epicentre to Tokyo have significantly affected occurrance of burst on Twitter.

リンク情報
DOI
https://doi.org/10.15099/00000836
CiNii Articles
http://ci.nii.ac.jp/naid/120005477887
CiNii Books
http://ci.nii.ac.jp/ncid/AA1214873X
URL
http://hdl.handle.net/2241/00122172
ID情報
  • DOI : 10.15099/00000836
  • ISSN : 1881-0101
  • CiNii Articles ID : 120005477887
  • CiNii Books ID : AA1214873X

エクスポート
BibTeX RIS