Topics and Influential User Identification in Twitter using Twitter Lists

• Item type: SIG Technical Reports(1)
• 公開日: 2014-07-25
• タイトル: Topics and Influential User Identification in Twitter using Twitter Lists
• タイトル / 言語: en
• タイトル: Topics and Influential User Identification in Twitter using Twitter Lists
• 言語: eng
• キーワード / 主題Scheme: Other
• キーワード / 主題: ソーシャルメディア
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_18gh
• 資源タイプ: technical report
• 著者所属: Waseda University
• 著者所属: Waseda University
• 著者所属: Waseda University／National Institute of Informatics
• 著者所属(英): en
• 著者所属(英): Waseda University
• 著者所属(英): en
• 著者所属(英): Waseda University
• 著者所属(英): en
• 著者所属(英): Waseda University / National Institute of Informatics
• 著者名: Guanying, Zhou; Hiroki, Asai; Hayato, Yamana
• 著者名(英): Guanying, Zhou; Hiroki, Asai; Hayato, Yamana
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: Twitter, as one of the most popular social network services, draws the attention of more and more researchers worldwide. With a large amount of information tweeted every day, it turns essential to identify the influential users we are interested in. In the previous research, researchers mainly identify topics from tweets and rank users by utilizing the follow relationship; however, the following relationship is strongly related to their reputation in real world and cannot describe their influence and activity level in Twitter exactly. Instead, in this paper, to identify topics and influential users, we use “Twitter List,” whose name represents the topic of listed members. By analyzing Twitter List, we are able to detect topics and identify influential users in the corresponding topic more efficiently. Based on our experimental evaluation using the selected two topics, the influential users identified by our proposed method have the average influence score related to the topic made by interviewees of 3.7 and 3.33 outweigh the methods of ranking by follower numbers with the average score of 3.22 and 3.27 respectively.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: Twitter, as one of the most popular social network services, draws the attention of more and more researchers worldwide. With a large amount of information tweeted every day, it turns essential to identify the influential users we are interested in. In the previous research, researchers mainly identify topics from tweets and rank users by utilizing the follow relationship; however, the following relationship is strongly related to their reputation in real world and cannot describe their influence and activity level in Twitter exactly. Instead, in this paper, to identify topics and influential users, we use “Twitter List,” whose name represents the topic of listed members. By analyzing Twitter List, we are able to detect topics and identify influential users in the corresponding topic more efficiently. Based on our experimental evaluation using the selected two topics, the influential users identified by our proposed method have the average influence score related to the topic made by interviewees of 3.7 and 3.33 outweigh the methods of ranking by follower numbers with the average score of 3.22 and 3.27 respectively.
• 書誌レコードID / 収録物識別子タイプ: NCID
• 書誌レコードID / 収録物識別子: AN10114171
• 書誌情報: 研究報告情報基礎とアクセス技術（IFAT）; 巻 2014-IFAT-115, 号 13, p. 1-6, 発行日 2014-07-25
• Notice: SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc.
• 出版者 / 言語: ja
• 出版者: 情報処理学会