Abstract
Most applications of Reinforcement Learning (RL) for dialogue have focused on slot-filling tasks. In this paper, we explore a task that requires negotiation, in which conversants need to exchange information in order to decide on a good solution. We investigate what information should be included in the system's RL state so that an optimal policy can be learned and so that the state space stays reasonable in size. We propose keeping track of the decisions that the system has made, and using them to constrain the system's future behavior in the dialogue. In this way, we can compositionally represent the strategy that the system is employing. We show that this approach is able to learn a good policy for the task. This work is a first step to a more general exploration of applying RL to negotiation dialogues.
Original language | English (US) |
---|---|
Title of host publication | Proceedings of the 2009 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2009 |
Pages | 450-455 |
Number of pages | 6 |
DOIs | |
State | Published - 2009 |
Event | 2009 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2009 - Merano, Italy Duration: Dec 13 2009 → Dec 17 2009 |
Other
Other | 2009 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2009 |
---|---|
Country | Italy |
City | Merano |
Period | 12/13/09 → 12/17/09 |
Fingerprint
ASJC Scopus subject areas
- Computer Vision and Pattern Recognition
- Human-Computer Interaction
- Signal Processing
Cite this
Representing the reinforcement learning state in a negotiation dialogue. / Heeman, Peter.
Proceedings of the 2009 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2009. 2009. p. 450-455 5373413.Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Representing the reinforcement learning state in a negotiation dialogue
AU - Heeman, Peter
PY - 2009
Y1 - 2009
N2 - Most applications of Reinforcement Learning (RL) for dialogue have focused on slot-filling tasks. In this paper, we explore a task that requires negotiation, in which conversants need to exchange information in order to decide on a good solution. We investigate what information should be included in the system's RL state so that an optimal policy can be learned and so that the state space stays reasonable in size. We propose keeping track of the decisions that the system has made, and using them to constrain the system's future behavior in the dialogue. In this way, we can compositionally represent the strategy that the system is employing. We show that this approach is able to learn a good policy for the task. This work is a first step to a more general exploration of applying RL to negotiation dialogues.
AB - Most applications of Reinforcement Learning (RL) for dialogue have focused on slot-filling tasks. In this paper, we explore a task that requires negotiation, in which conversants need to exchange information in order to decide on a good solution. We investigate what information should be included in the system's RL state so that an optimal policy can be learned and so that the state space stays reasonable in size. We propose keeping track of the decisions that the system has made, and using them to constrain the system's future behavior in the dialogue. In this way, we can compositionally represent the strategy that the system is employing. We show that this approach is able to learn a good policy for the task. This work is a first step to a more general exploration of applying RL to negotiation dialogues.
UR - http://www.scopus.com/inward/record.url?scp=77949374322&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=77949374322&partnerID=8YFLogxK
U2 - 10.1109/ASRU.2009.5373413
DO - 10.1109/ASRU.2009.5373413
M3 - Conference contribution
AN - SCOPUS:77949374322
SN - 9781424454792
SP - 450
EP - 455
BT - Proceedings of the 2009 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2009
ER -