Virtual Agent Command Disambiguation via Ontology Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Virtual agents face challenges in interpreting generic commands due to a lack of specificity, often referencing multiple objects, which hinders their ability to perform desired actions accurately.
Innovation Solution
The implementation of an ontology tree that assigns user and physical objects, allowing for disambiguation of generic commands by analyzing historical associations, recent interactions, and object states, enabling the selection of a target object based on weighted scores.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If generic commands are used for controlling objects, then ease of operation is improved, but measurement precision deteriorates because the command cannot uniquely identify a target object
Solution Approach 1:
The system uses feedback from historical associations and recent interactions to resolve ambiguity in generic commands. By analyzing past usage patterns and current context, the virtual agent determines the most likely target object without requiring users to provide specific identifiers, thus maintaining ease of operation while improving identification accuracy.
Solution Approach 2:
The system performs preliminary analysis of candidate objects based on historical associations and recent interactions before executing the command. This pre-processing step ranks potential target objects, allowing the system to quickly select the most probable target and resolve ambiguity before the actual action is performed.
2Measurement precision
If specific object identifiers are required for commands, then target object identification accuracy is improved, but ease of operation deteriorates because users must know and remember specific identifiers
Solution Approach 1:
The system performs self-service by automatically resolving object ambiguity using its own historical data and context information. Instead of requiring users to provide specific identifiers, the virtual agent independently analyzes candidate objects and selects the most likely target based on historical associations and recent interactions, making the system self-sufficient in disambiguation tasks.
3Measurement precision
If multiple candidate objects are considered for disambiguation, then target object identification accuracy is improved, but device complexity increases due to the need to analyze historical associations and recent interactions
Solution Approach 1:
The system achieves multi-functionality by using a single disambiguation mechanism that handles multiple candidate objects through unified analysis of historical associations and recent interactions. This universal approach allows the system to process various types of objects and commands through the same framework, reducing overall system complexity despite handling multiple candidates.
Data Source
AI summary
A computer system performs an action on an object identified from a command. A command is analyzed to perform an action on a target object, wherein the command includes a term for the target object that refers to a plurality of different candidate objects. The target object is identified from the plurality of different candidate objects based on historical associations of the term with specific ones of the candidate objects, recent interactions with the different candidate objects, and a state of the candidate objects provided by network devices. The action is performed on the identified target object. Embodiments of the present invention further include a method and program product for performing an action on an object identified from a command in substantially the same manner described above.