Voice Agent Task Sequence Optimization via Dynamic Step Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing voice recognition agent services on mobile electronic devices often perform application functions through unnecessary steps, wasting time and power due to the inclusion of non-essential steps in user-defined task sequences.
Innovation Solution
An electronic apparatus that analyzes user utterances and motions, applying weights to step information to optimize task sequences by excluding unnecessary steps based on weight thresholds and task performance, updating task information dynamically to execute voice recognition agent functions efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a voice recognition agent performs application functions through a plurality of steps set by a user, then the function can be executed, but unnecessary steps are included which waste time and power
Solution Approach 1:
The patent extracts and removes unnecessary steps from the task sequence by analyzing user utterances and motions. The voice recognition agent identifies steps that are not essential for achieving the task goal and excludes them from execution, thereby reducing time consumption while maintaining execution efficiency.
Solution Approach 2:
The task sequence is made dynamic and adaptable based on real-time user input. The system dynamically adjusts the execution plan by reweighting steps based on user utterances and motions, allowing the agent to optimize the task sequence during execution rather than following a fixed predetermined sequence.
2Productivity
If a voice recognition agent performs application functions through a plurality of steps set by a user, then the function can be executed, but unnecessary steps are included which waste power
Solution Approach 1:
The patent extracts and removes unnecessary steps from the task sequence by analyzing user utterances and motions. By eliminating steps that do not contribute to the task goal, the system reduces the computational and operational overhead, thereby lowering power consumption while maintaining execution efficiency.
Solution Approach 2:
The system performs only the necessary portion of the task sequence rather than executing all predetermined steps. By applying partial action - executing only essential steps identified through user input analysis - the system avoids excessive power consumption associated with unnecessary operations.
3Loss of time
If the voice recognition agent excludes steps from the task sequence, then execution time is reduced, but task performance may be affected
Solution Approach 1:
The system incorporates feedback mechanisms where user utterances and motions continuously inform the task execution process. The voice recognition agent monitors user input in real-time and adjusts step inclusion/exclusion decisions based on feedback, ensuring that task performance is maintained while minimizing time consumption.
Solution Approach 2:
The task sequence is made dynamic and adaptable based on real-time user input. The system dynamically adjusts the execution plan by reweighting steps based on user utterances and motions, allowing the agent to optimize the task sequence during execution rather than following a fixed predetermined sequence.
Data Source
AI summary
An electronic apparatus is disclosed. The electronic apparatus includes: a memory storing task information and keyword information regarding a voice recognition agent function related to execution of an application, and a processor connected with the memory and configured to control the electronic apparatus. The processor may be configured to: apply a first weight to first step information including the keyword information among a plurality of step information included in the task information, apply a second weight different from the first weight to second step information including a motion among the plurality of step information, update the task information based on the first step information to which the first weight was applied and the second step information to which the second weight was applied, and based on a voice command for the application being input, execute the voice recognition agent function based on the updated task information.