Privacy Preserving LAMA. Maliah, S., Shani, G., & Stern, R. In
Privacy Preserving LAMA [link]Paper  abstract   bibtex   
In collaborative privacy preserving planning (CPPP), multiple agents collaborate to achieve a goal while keeping certain facts about the world private. A prominent approach in the development of CPPP algorithms is to use components from single agent planners and adapt them to preserve privacy. In this short paper, we show how the components of LAMA, arguably one of the most successful single-agent planners, can be used in a privacy preserving manner. These components include alternating between a landmark heuristic and an FF heuristic, preferred operators and deferred heuristic evaluation. We integrate the components into the Greedy Privacy Preserving Planner, a state-of-the-art CPPP algorithm. The resulting algorithm performs better than other CPPP algorithms from the recent Competition of Distributed and Multiagent Planners.

Downloads: 0