I've said it before, and I'll say it again: an empirical investigation of the upper bound of the selection approach to dialogue

Authors:
Sudeep Gandhe;David Traum
Affiliations:
Institute for Creative Technologies, Fiji way, Marina del Rey, CA;Institute for Creative Technologies, Fiji way, Marina del Rey, CA
Venue:
SIGDIAL '10 Proceedings of the 11th Annual Meeting of the Special Interest Group on Discourse and Dialogue
Year:
2010

Citing 9
Cited 1

The TRAINS 93 Dialogues

The TRAINS 93 Dialogues
The Monroe Corpus

The Monroe Corpus
A corpus-based approach to help-desk response generation

CIMCA '06 Proceedings of the International Conference on Computational Inteligence for Modelling Control and Automation and International Conference on Intelligent Agents Web Technologies and International Commerce
Virtual Patients for Clinical Therapist Skills Training

IVA '07 Proceedings of the 7th international conference on Intelligent Virtual Agents
Semi-formal Evaluation of Conversational Characters

Languages: From Formal to Natural
An evaluation understudy for dialogue coherence models

SIGdial '08 Proceedings of the 9th SIGdial Workshop on Discourse and Dialogue
Building effective question answering characters

SigDIAL '06 Proceedings of the 7th SIGdial Workshop on Discourse and Dialogue
The Meteor metric for automatic evaluation of machine translation

Machine Translation
SWITCHBOARD: telephone speech corpus for research and development

ICASSP'92 Proceedings of the 1992 IEEE international conference on Acoustics, speech and signal processing - Volume 1

Toward learning and evaluation of dialogue policies with text examples

SIGDIAL '11 Proceedings of the SIGDIAL 2011 Conference

Quantified Score

Hi-index	0.00

Visualization

Abstract

We perform a study of existing dialogue corpora to establish the theoretical maximum performance of the selection approach to simulating human dialogue behavior in unseen dialogues. This maximum is the proportion of test utterances for which an exact or approximate match exists in the corresponding training corpus. The results indicate that some domains seem quite suitable for a corpus-based selection approach, with over half of the test utterances having been seen before in the corpus, while other domains show much more novelty compared to previous dialogues.