Dynamic Response Correction for User Inquiry Intent Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing inquiry responding systems using AI models struggle to provide accurate responses to user inquiries due to fixed response messages, leading to user dissatisfaction and perceived misrecognition of inquiry intents.
Innovation Solution
An electronic device processes user inquiries using a natural language understanding (NLU) model to recognize intent, identify a representative inquiry, compare embedding vectors, extract keywords, and correct response messages dynamically based on the extracted keywords.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a fixed response message is provided for an inquiry intent, then the system operation is simple, but the response accuracy deteriorates
Solution Approach 1:
The patent applies dynamics by transitioning from fixed response messages to dynamic response generation. The system now adapts response messages based on extracted keywords and their importance values, allowing the response to change according to the specific inquiry content while maintaining systematic operation through automated keyword-based selection processes.
Solution Approach 2:
The patent changes the parameter of response message selection from static (fixed per intent) to dynamic (based on keyword importance values). By calculating importance values for extracted keywords and selecting responses based on these varying parameters, the system achieves both operational simplicity through automation and improved response accuracy through content-based adaptation.
2Productivity
If various inquiry contents are classified as having the same inquiry intent, then the classification efficiency is improved, but the user satisfaction deteriorates
Solution Approach 1:
The patent applies local quality by differentiating the treatment of inquiries within the same intent category. Instead of applying a uniform fixed response to all inquiries of a given intent, the system extracts specific keywords from each inquiry, calculates their importance values, and selects or generates responses tailored to the local characteristics of each inquiry content, thereby maintaining classification efficiency while improving user satisfaction.
Solution Approach 2:
The patent segments the response generation process by first classifying inquiries into intents (maintaining efficiency) and then further segmenting based on extracted keywords and their importance values. This two-stage segmentation allows the system to handle multiple inquiries efficiently through intent classification while providing differentiated, accurate responses through keyword-based segmentation within each intent category.
3Device complexity
If a fixed response message is used, then the device complexity is reduced, but the information accuracy deteriorates
Solution Approach 1:
The patent applies self-service by enabling the system to automatically extract keywords, calculate their importance values, and select or generate appropriate response messages without human intervention. This automated self-service mechanism reduces the need for complex manual configuration of fixed responses for each possible inquiry variation while maintaining high information accuracy through dynamic, content-based response selection.
Solution Approach 2:
The patent applies preliminary action by pre-processing inquiries to extract keywords and calculate their importance values before response generation. This preliminary analysis enables the system to prepare the necessary information for accurate response selection in advance, reducing the complexity of real-time decision-making while ensuring information accuracy through proactive keyword extraction and importance assessment.
Data Source
AI summary
An electronic device for providing a corrected response message according to an utterance intention of a user by using a keyword included in an input inquiry input by the user, and an operation method of the electronic device are provided. The electronic device includes receiving the input inquiry input by the user, identifying a representative inquiry according to an utterance intention of the user by analyzing the input inquiry by using a natural language understanding (NLU) model, extracting a keyword from the input inquiry by comparing a vector value of a first embedding vector of the input inquiry changed through the NLU model with a vector value of a second embedding vector of the representative inquiry, and correcting a response message mapped to correspond to the representative inquiry, by using the extracted keyword.