Digital Assistant Confirmation Control for Ambiguous Responses
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electronic devices struggle to effectively manage user input and output in digital assistant interactions, particularly when user responses need confirmation or prudent decision-making is required, leading to potential misalignment with third-party opinions or prolonged response times.
Innovation Solution
An electronic device equipped with a digital assistant that processes user utterances, generates appropriate responses, and identifies situations requiring user confirmation, using learning models and natural language processing to manage interactions and provide notifications for user intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If the digital assistant automatically generates responses without user confirmation, then response speed is improved, but accuracy and alignment with user intent deteriorate
Solution Approach 1:
The system dynamically adjusts the level of automation based on the situation. For routine queries, it responds automatically without confirmation (high speed). For complex or ambiguous queries, it switches to confirmation mode (lower speed but higher accuracy). This dynamic switching resolves the contradiction by adapting the response generation mode to the specific context.
Solution Approach 2:
The system implements feedback loops where user responses are analyzed to determine whether confirmation is needed. The digital assistant learns from user feedback patterns and adjusts its confirmation requests accordingly. This feedback mechanism ensures that automatic responses are generated only when confidence is high, maintaining both speed and accuracy.
2Measurement precision
If the digital assistant requests user confirmation for every response, then accuracy is improved, but response time increases
Solution Approach 1:
Instead of requesting confirmation for every response (excessive action), the system applies confirmation only partially - specifically for queries that meet certain criteria such as complexity, ambiguity, or importance. This selective approach maintains accuracy where needed while avoiding unnecessary delays for routine queries.
Solution Approach 2:
The system changes the parameter of confirmation frequency based on query characteristics. For simple, clear queries, the confirmation parameter is set to false (no confirmation). For complex or ambiguous queries, the confirmation parameter is set to true. This parameter change strategy resolves the time-accuracy tradeoff by adapting confirmation behavior to query specifics.
3Measurement precision
If the digital assistant processes all user inputs through complex learning models, then response accuracy is improved, but computational complexity and processing time increase
Solution Approach 1:
The processing system is segmented into multiple levels: a first learning model handles simple, routine queries with low computational complexity, while a second, more complex learning model handles only the subset of queries that require higher accuracy. This segmentation reduces overall computational complexity while maintaining accuracy where needed.
Solution Approach 2:
The system performs preliminary analysis of user inputs to categorize them by complexity and required accuracy level before routing to appropriate learning models. This preliminary action allows the system to avoid unnecessary complex processing for simple queries, reducing computational overhead while maintaining response accuracy for complex queries.
Data Source
Figure 1
Figure 2-a~2-b
Figure 3-a~3-c
AI summary
Disclosed is an electronic device. The electronic device may execute an application for transmitting and receiving at least one of text data or voice data with another electronic device using the communication module, in response to occurrence of at least one event, based on receiving at least one of text data or voice data from the another electronic device, identify that a confirmation is necessary using the digital assistant based on at least one of text data or voice data being generated based on a characteristic of ab utterance using a digital assistant, generate a notification to request confirmation using the digital assistant based on confirmation being necessary, and output the notification using the application. A method for identifying that a confirmation is necessary may include identifying using voice data or text data that is received from another electronic device using a rule-based or AI algorithm. When a confirmation is necessary is identified using the AI algorithm, the method may use machine learning, neural network, or a deep learning algorithm.