Toward Natural Turn-Taking in a Virtual Human Negotiation Agent

Abstract

In this paper we assess our progress toward creating avirtual human negotiation agent with fluid turn-takingskills. To facilitate the design of this agent, we have collected a corpus of human-human negotiation roleplaysas well as a corpus of Wizard-controlled human-agentnegotiations in the same roleplay scenario. We comparethe natural turn-taking behavior in our human-humancorpus with that achieved in our Wizard-of-Oz corpus,and quantify our virtual humans turn-taking skills using a combination of subjective and objective metrics.We also discuss our design for a Wizard user interfaceto support real-time control of the virtual humans turntaking and dialogue behavior, and analyze our wizardsusage of this interface.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 2015
Accession Number
AD1159900

Entities

People

  • David Devault
  • Johnathan Mell
  • Jonathan Gratch

Organizations

  • University of Southern California

Tags

Communities of Interest

  • Human Systems

DTIC Thesaurus Topics

  • Agreements
  • Bargaining
  • Computers
  • Control Systems
  • Data Sets
  • Dialogue Systems
  • Engineering
  • Language
  • Models
  • Multiagent Systems
  • Negotiations
  • New York
  • Psychology
  • Social Psychology
  • United States
  • United States Government
  • User Interface

Fields of Study

  • Computer science

Readers

  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.