Output details
11 - Computer Science and Informatics
Heriot-Watt University
Article title
Hybrid reinforcement/supervised learning of dialogue policies from fixed data sets
Type
D - Journal article
Title of journal
Computational Linguistics
Article number
-
Volume number
34
Issue number
4
First page of article
487
ISSN of journal
0891-2017
Year of publication
2008
URL
-
Number of additional authors
2
Additional information
<20>This work became the basis of an EC FP7 project "CLASSiC" (Computational Learning in Adaptive Systems for Spoken Conversation, 5m euros, 2008-2011), coordinated by Lemon, which developed an end-to-end statistical approach to Spoken Dialogue Systems. This then led to EC project "PARLANCE" (FP7, 2011,1m euro), which extends these methods to incremental dialogue. The work also generated an ESPRC project (£358,000, 2009), with Lemon as PI, exploring state compression methods for complex extensions of the problem. This research also led to special sessions at Interspeech 2009 and 2011 organised by Lemon, and an invited talk at ECAI 2012.
Interdisciplinary
-
Cross-referral requested
-
Research group
B - Interaction
Citation count
36
Proposed double-weighted
No
Double-weighted statement
-
Reserve for a double-weighted output
No
Non-English
No
English abstract
-
