Do you get it? user-evaluated explainable BDI agents

Authors:
Joost Broekens;Maaike Harbers;Koen Hindriks;Karel Van Den Bosch;Catholijn Jonker;John-Jules Meyer
Affiliations:
Delft University of Technology;Utrecht University;Delft University of Technology;TNO Institute of Defence, Security and Safety, The Netherlands;Delft University of Technology;Utrecht University
Venue:
MATES'10 Proceedings of the 8th German conference on Multiagent system technologies
Year:
2010

Citing 7
Cited 1

Agents that learn to explain themselves

AAAI'94 Proceedings of the twelfth national conference on Artificial intelligence (vol. 2)
Character-Based Interactive Storytelling

IEEE Intelligent Systems
Teaching Negotiation Skills through Practice and Reflection with Virtual Humans

Simulation
Building explainable artificial intelligence systems

IAAI'06 Proceedings of the 18th conference on Innovative applications of artificial intelligence - Volume 2
An explainable artificial intelligence system for small-unit tactical behavior

IAAI'04 Proceedings of the 16th conference on Innovative applications of artifical intelligence
A Study into Preferred Explanations of Virtual Agent Behavior

IVA '09 Proceedings of the 9th International Conference on Intelligent Virtual Agents
AutoTutor: an intelligent tutoring system with mixed-initiative dialogue

IEEE Transactions on Education

Virtual reality negotiation training increases negotiation knowledge and skill

IVA'12 Proceedings of the 12th international conference on Intelligent Virtual Agents

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper we focus on explaining to humans the behavior of autonomous agents, i.e., explainable agents. Explainable agents are useful for many reasons including scenario-based training (e.g. disaster training), tutor and pedagogical systems, agent development and debugging, gaming, and interactive storytelling. As the aim is to generate for humans plausible and insightful explanations, user evaluation of different explanations is essential. In this paper we test the hypothesis that different explanation types are needed to explain different types of actions. We present three different, generically applicable, algorithms that automatically generate different types of explanations for actions of BDI-based agents. Quantitative analysis of a user experiment (n=30), in which users rated the usefulness and naturalness of each explanation type for different agent actions, supports our hypothesis. In addition, we present feedback from the users about how they would explain the actions themselves. Finally, we hypothesize guidelines relevant for the development of explainable BDI agents.