You are in : Home » Results & submissions » Select UOA » 11 - Computer Science and Informatics » View submission: Heriot-Watt University » Outputs » Detail

Output details

11 - Computer Science and Informatics

Heriot-Watt University

Return to search Previous output Next output

Output 42 of 91 in the submission

Article title

Learning and evaluation of dialogue strategies for new applications : Empirical methods for optimization from small data sets

Type

D - Journal article

DOI

10.1162/coli_a_00038

Title of journal

Computational Linguistics

Article number

Volume number

Issue number

First page of article

153

ISSN of journal

0891-2017

Year of publication

2011

URL

Number of additional authors

Additional information

<20>This work extends an academically acclaimed ACL 2008 paper (approx. 20% acceptance rate) and is the first to employ a fully data-driven framework to optimise and evaluate Reinforcement Learning-based interaction strategies. This formed the basis of several work-packages applying and extending the techniques in 4 EC-funded projects managed by Lemon: SpaceBook (FP7, 645K euro, 2011), JAMES (FP7, 741K euro, 2011), Help4Mood (FP7, 290K euro, 2010), STAC (ERC, 437K euro, 2011). A book based on this work has sold over 750 copies since publication in November 2011. Lemon gave an invited talk on this research at Nuance Communications, California, 2013.

Interdisciplinary

Cross-referral requested

Research group

B - Interaction

Citation count

Proposed double-weighted

Double-weighted statement

Reserve for a double-weighted output

Non-English

English abstract