2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07 2007
DOI: 10.1109/icassp.2007.367185
|View full text |Cite
|
Sign up to set email alerts
|

The Hidden Information State Approach to Dialog Management

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
78
0

Year Published

2008
2008
2014
2014

Publication Types

Select...
4
2
1

Relationship

2
5

Authors

Journals

citations
Cited by 95 publications
(78 citation statements)
references
References 7 publications
0
78
0
Order By: Relevance
“…The most common acts are listed in Table 1 and a simple dialogue illustrating their use is shown in Table 2. A full description of the dialogue act set used by the HIS system is given in Young (2007).…”
Section: Dialogue Actsmentioning
confidence: 99%
See 1 more Smart Citation
“…The most common acts are listed in Table 1 and a simple dialogue illustrating their use is shown in Table 2. A full description of the dialogue act set used by the HIS system is given in Young (2007).…”
Section: Dialogue Actsmentioning
confidence: 99%
“…The system typically maintains a distribution of upto several hundred partitions corresponding to many thousands of dialogue states. The HIS system has been described in outline in a number of conference papers (Young, 2006;Young et al, 2007;. The aim of this paper is to provide a single coherent and more detailed description of how the HIS system works, and an assessment of its performance characteristics.…”
mentioning
confidence: 99%
“…This prior information is included in the learning process by grouping such actions together. The groups are called summary actions and play a similar role to the summary actions of the HIS model [3]. Typical summary actions will be to "request", "confirm" the most likely value or to "select" between the top two values for a particular slot.…”
Section: Learning With Budsmentioning
confidence: 99%
“…This number grows exponentially with the number of slots in the system. In the Hidden Information State model, learning was implemented by summarising the full belief distribution according to the beliefs in the most likely partitions [3]. An algorithm for online learning of the Q function was then implemented using a grid-based approximation.…”
Section: Policy Designmentioning
confidence: 99%
See 1 more Smart Citation