Cognitive Model of Response Token Generation

DOI Web Site

Bibliographic Information

Other Title
  • 相槌生成の認知的モデル
  • アイズチ セイセイ ノ ニンチテキ モデル

Search this article

Abstract

<p>In this study, we propose a cognitive model of response token (RT) generation. First, we conducted a quantitative analysis about the occurrence position of RTs with respect to TCU. As a result, it was found that responsive interjections are frequently used within TCU, while other forms of RTs are used at the end of TCU. Second, we conducted a qualitative analysis when and how RTs are used in the long turn. As a result, we found that some responsive interjections are used in positions where information is incomplete and other forms of RTs are used in positions where information is complete or predictable. Based on these results, we finally propose two parallel processes of RT generation.</p>

Journal

Details 詳細情報について

Report a problem

Back to top