Agent Response Selection Without Repeating Intent Inference

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing agent systems struggle to accurately infer the intent of user questions when multiple responses are presented, leading to situations where the intent cannot be inferred from selection results, particularly when users respond with numbers or symbols.

Innovation Solution

An agent device with a processor that receives user input, executes inference processing to determine the intent of questions, provides option information when multiple responses are acquired, and determines whether new input is selection information, thereby avoiding unnecessary inference processing by directly providing response information associated with selection results.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the agent executes inference processing on all user inputs to determine intent, then the precision of intent inference is improved, but the processing load increases and response time is delayed

Engineering Contradiction:
Improveprecision of intent inferenceVSAvoidprocessing speed
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent segments user inputs into two categories: selection information (requiring no inference) and question information (requiring inference). This segmentation allows the system to apply inference processing only where necessary, reducing overall processing load while maintaining inference precision for actual questions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies partial action by executing inference processing only for a portion of inputs (questions) rather than all inputs. Selection information is processed directly without inference, achieving partial processing that reduces computational overhead while maintaining accuracy for cases that require it.

Inventive Principle:
Principle #16Partial or excessive action

2Stability of the object's composition

If the agent executes inference processing on selection information, then the system maintains consistent processing logic, but the processing load increases unnecessarily and response time is delayed

Engineering Contradiction:
Improveconsistency of processing logicVSAvoidresponse time
Core Design Contradiction:
Stability of the object's compositionVSLoss of time

Solution Approach 1:

The patent segments processing paths based on input type: selection information follows a direct retrieval path, while question information follows an inference path. This segmentation maintains logical consistency by treating different input types appropriately while avoiding unnecessary processing steps.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary classification of input information to determine whether it requires inference processing. By identifying selection information in advance, the system can bypass inference processing and directly retrieve associated responses, reducing response time while maintaining logical consistency.

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If the agent provides multiple responses as options, then the user can select from accurate possibilities, but the system cannot infer intent from subsequent selection results

Engineering Contradiction:
Improveaccuracy of response optionsVSAvoidloss of intent information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent implements a feedback mechanism where the system provides multiple response options and receives user selections as feedback. Instead of attempting to infer intent from the selection (which fails), the system uses the selection to directly retrieve and provide the associated response, effectively utilizing the feedback without requiring intent inference.

Inventive Principle:
Principle #23Feedback

Data Source

PatentUS11809885B2Agent device, agent system, and recording medium
Publication Date: 2023.11.07 TOYOTA JIDOSHA KK
  • US11809885B2 patent drawing
  • US11809885B2 patent drawing
  • US11809885B2 patent drawing

AI summary

An agent device receives input information that is input by the user, in a case in which the input information is a question from the user, executes inference processing on the input information to infer an intent of the question in order to acquire a response to the question based on the intent, in a case in which a plurality of the responses are acquired, provides the notification device with option information that includes the plurality of responses as options, in a case in which new input information is received, determines whether the new input information is information requiring the inference processing or is selection information relating to a selection result from selection of the options, and in a case in which the new input information is the selection information, provides the notification device with response information regarding the response associated with the selection result without executing the inference processing.