Guided Plan Recognition Observation Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current AI plan recognition systems face inefficiencies due to the lack of guidance in selecting observations, leading to inconsistent or incompatible outcomes, which results in wasted computing resources and unsatisfactory results.
Innovation Solution
The implementation of a guided plan recognition method that involves obtaining a specification of a plan recognition problem, formulating planning problems, solving them to determine relevant plans, and post-processing to select appropriate observations for improved performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual observation selection is used in plan recognition systems, then system simplicity is maintained, but computing resources are wasted due to selection of inappropriate observations and multiple unsatisfactory runs
Solution Approach 1:
The system performs preliminary analysis by formulating planning problems and solving them to determine relevant plans before final observation selection. This preliminary action identifies which observations are most valuable, preventing wasted computing resources on inappropriate observations later in the process.
Solution Approach 2:
The system introduces an intermediary planning module that acts as a mediator between the observation database and the plan recognition process. This intermediary formulates and solves planning problems to determine which observations should be selected, bridging the gap between raw observations and effective plan recognition.
2Reliability
If guidance is added to observation selection, then observation quality and consistency are improved, but system complexity increases
Solution Approach 1:
The system implements feedback by solving planning problems and using the results to guide observation selection. The planning problem solutions provide feedback on which observations are relevant and consistent, ensuring that selected observations can be explained together and improving overall observation quality.
Solution Approach 2:
The system performs preliminary planning problem formulation and solution before final observation selection. This preliminary action ensures that observations are pre-screened for consistency and relevance, improving reliability before the actual plan recognition process begins.
3Measurement precision
If multiple planning problems are solved to determine relevant plans, then observation selection accuracy is improved, but computational time increases
Solution Approach 1:
The system segments the overall plan recognition task into multiple planning problems that can be solved independently. By dividing the problem into smaller, manageable planning tasks, the system can determine relevant plans more efficiently while maintaining high observation selection accuracy.
Solution Approach 2:
The system solves planning problems as a preliminary step to identify relevant plans before final observation selection. This preliminary solving of planning problems provides accurate guidance for observation selection, reducing the need for multiple trial runs and actually saving computational time in the long run.
Data Source
AI summary
Performance of a computer running a plan recognition application is improved by obtaining, with a user interface implemented on the computer, a specification of a plan recognition problem, including a plurality of candidate observations; formulating at least one planning problem, with the computer, based on the specification; solving the at least one planning problem, with the computer, to determine at least one plan. The at least one plan is post-processed, with the computer, to determine at least one of the candidate observations which should be selected to solve the plan recognition problem; and the plan recognition problem is solved, with the computer, using the at least one of the candidate observations which should be selected to solve the plan recognition problem. Less CPU time is typically required for the solution as compared to techniques without guidance for selecting the observations.