グループディスカッションコーパスの構築および性格特性との関連性の分析  [in Japanese] Development of Group Discussion Interaction Corpus and Analysis of the Relationship with Personality Traits  [in Japanese]

Access this Article

Search this Article

Author(s)

Abstract

我々は,グループディスカッションを題材とし,グループ参加者に見られるマルチモーダル情報に基づき参加者のコミュニケーションスキルを測定・評価することを目的としたプロジェクトに取り組んでいる.本研究では,コミュニケーションスキル分析のために重要な資源となる,外部専門家によるコミュニケーションスキル評価値および,参加者のBig-five性格特性を含む,マルチモーダル会話コーパスを構築することを目的とする.本論文では,コミュニケーションスキルに関連する特性を示すための初期分析として,Big-five性格特性に見られる発話の韻律特徴および,頭部加速度センサの加速度変化量について多項ロジスティック回帰分析を行い,その判別モデルの推定結果を議論する.Our research project tackles analyzing and evaluating communication skills based on multimodal information in group discussion situations. In order to propose the data resources toward analyzing communication skills, this research conducts data collection experiments to construct group discussion corpus which stores participants' speech, gaze, head motions, and poses using some multimodal measurement devices. The corpus includes participants' evaluated values of communication skills by external experts, and their Big-five personality traits scores for the communication skills analysis. Based on the multimodal corpus, we discuss the relationship of the participants' prosody and variation of acceleration sensor features with the Big-five personality traits using multinomial logistic regression analysis.

Our research project tackles analyzing and evaluating communication skills based on multimodal information in group discussion situations. In order to propose the data resources toward analyzing communication skills, this research conducts data collection experiments to construct group discussion corpus which stores participants' speech, gaze, head motions, and poses using some multimodal measurement devices. The corpus includes participants' evaluated values of communication skills by external experts, and their Big-five personality traits scores for the communication skills analysis. Based on the multimodal corpus, we discuss the relationship of the participants' prosody and variation of acceleration sensor features with the Big-five personality traits using multinomial logistic regression analysis.

Journal

  • IPSJ Journal

    IPSJ Journal 56(4), 1217-1227, 2015-04-15

    Information Processing Society of Japan (IPSJ)

Codes

  • NII Article ID (NAID)
    110009890361
  • NII NACSIS-CAT ID (NCID)
    AN00116647
  • Text Lang
    JPN
  • Article Type
    Journal Article
  • ISSN
    1882-7764
  • Data Source
    NII-ELS  IPSJ 
Page Top