首页    期刊浏览 2024年11月28日 星期四
登录注册

文章基本信息

  • 标题:Predicting Turn Taking from Gaze Transition Patterns Considering Participation Status in Multi-Party Conversation
  • 本地全文:下载
  • 作者:Takashi Makino ; Yoshinari Takegawa ; Keiji Hirata
  • 期刊名称:Lecture Notes in Engineering and Computer Science
  • 印刷版ISSN:2078-0958
  • 电子版ISSN:2078-0966
  • 出版年度:2018
  • 卷号:2237&2238
  • 页码:600-605
  • 出版社:Newswood and International Association of Engineers
  • 摘要:In multi-party conversation, listener’s role is not uniform. Listeners are divided into those who are addressed by the speaker (addressee) and those who are not addressed by the speaker (side-participant). In this paper, we propose a mathematical model to predict whether turn-taking or turnkeeping occurs. For a feature quantity of a model, we focus on gaze transition patterns near the end of utterance for each role of a participant. We analyze the difference in frequencies of gaze transition patterns between addressee and sideparticipant. Based on the result of analysis, we construct a probabilistic model that considers listener’s roles. As a result, the proposed model outperforms the conventional model that does not consider listener’s roles.
  • 关键词:turn;taking; participation status; gaze; multiparty conversation;
国家哲学社会科学文献中心版权所有