Multi-agent plan recognition

1. A computer program product for solving a multi-agent plan recognition problem, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a planning projector component to cause the planning projector component to:

transform, by the planning projector component, a problem model and an at least partially ordered sequence of observations into an artificial intelligence planning problem through a transform algorithm, wherein the problem model comprises a domain description from a plurality of agents and a durative action and wherein at least one of the observations of the at least partially ordered sequence of observations is a condition that changes over time, wherein the transformation comprises:

compiling away observations from a simplified input data employing special predicates for fluents corresponding to the at least partially ordered sequence of observations in a manner that ensures that the order is preserved, and wherein a type of the compiling away is based on the nature of observations of the at least partially ordered sequence of observations; and

discarding an observation of the at least partially ordered sequence of observations based on a determination that the observation comprises an unreliable element;

determine, by the planning projector component, plan information using an artificial intelligence planner on the artificial intelligence planning problem, wherein the determining the plan information comprises assessing a penalty based on the discarding the observation, wherein the penalty comprises increasing a duration of an action associated with a discarded observation; and

translate, by the planning projector component, the plan information into information indicative of a solution to the artificial intelligence planning problem.

2. The computer program product of claim 1 , wherein the partially ordered sequence of observations includes an action condition, and wherein the program instructions further cause the planning projector component to:

translate the action condition into a second condition that changes over time.

3. The computer program product of claim 1 , wherein a description of the problem model comprises an attribute selected from the group consisting of one or more facts, an initial state, one or more action operators, a goal condition, and a number of agents in the problem model.

4. The computer program product of claim 1 , wherein a description of the problem model comprises a defined initial state.

5. The computer program product of claim 1 , wherein a description of the problem model comprises a defined goal condition.

6. The computer program product of claim 1 , wherein the at least partially ordered sequence of observations comprises a missing observation.

7. The computer program product of claim 1 , wherein the artificial intelligence planner comprises a planning component selected from the group consisting of a top-K planning component, a diverse planning component, a delta planning component, and a hybrid planning component.

8. A system, comprising:

a memory that stores computer executable components; and

a processor that executes the computer executable components stored in the memory, wherein the computer executable components comprise:

a planning projector component that:

transforms a problem model and an at least partially ordered sequence of observations into an artificial intelligence planning problem through a transform algorithm, wherein the problem model comprises a domain description from a plurality of agents and a durative action and wherein at least one of the observations of the at least partially ordered sequence of observations is a condition that changes over time, and wherein the transformation comprises:

compiling away observations from simplified input data employing special predicates for fluents corresponding to the at least partially ordered sequence of observations in a manner that ensures that the order is preserved; and

discarding an observation of the at least partially ordered sequence of observations based on a determination that the observation comprises an unreliable element;

determines plan information using an artificial intelligence planner on the artificial intelligence planning problem, wherein the determining the plan information comprises assessing a penalty based on the discarding the observation, wherein the penalty comprises increasing a duration of an action associated with a discarded observation; and

translates the plan information into information indicative of a solution to the artificial intelligence planning problem.

9. The system of claim 8 , wherein the artificial intelligence planner comprises a planning component selected from the group consisting of a top-K planning component, a diverse planning component, a delta planning component, and a hybrid planning component.

10. The system of claim 8 , wherein a type of the compiling away is based on the nature of observations of the at least partially ordered sequence of observations.

11. The system of claim 8 , wherein the penalty comprises a temporal penalty set by defining a duration to the discarding, that exceeds durations defined for other actions.

12. The computer program product of claim 1 , wherein the penalty comprises a temporal penalty.