Learning to infer final plans in human team planning
Author(s)
Kim, Joseph; Woicik, Matthew E.; Gombolay, Matthew C.; Son, Sung-Hyun; Shah, Julie A
DownloadAccepted version (453.4Kb)
Terms of use
Metadata
Show full item recordAbstract
We envision an intelligent agent that analyzes conversations during human team meetings in order to infer the team's plan, with the purpose of providing decision support to strengthen that plan. We present a novel learning technique to infer teams' final plans directly from a processed form of their planning conversation. Our method employs reinforcement learning to train a model that maps features of the discussed plan and patterns of dialogue exchange among participants to a final, agreed-upon plan. We employ planning domain models to efficiently search the large space of possible plans, and the costs of candidate plans serve as the reinforcement signal. We demonstrate that our technique successfully infers plans within a variety of challenging domains, with higher accuracy than prior art. With our domain-independent feature set, we empirically demonstrate that our model trained on one planning domain can be applied to successfully infer team plans within a novel planning domain.
Date issued
2018Department
Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory; Lincoln LaboratoryJournal
Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence
Publisher
International Joint Conferences on Artificial Intelligence
Citation
Kim, Joseph, et al., "Learning to infer final plans in human team planning." Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence, Stockholm, July 13-19, 2018, edited by Jérôme Lang, International Joint Conferences on Artificial Intelligence, 2018: p. 4771-79 doi 10.24963/IJCAI.2018/663 ©2018 Author(s)
Version: Author's final manuscript
ISBN
978-0-9992411-2-7