Query Masking for Trust-Based Secondary Digital Assistants
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automated assistants often require explicit invocation and may not process user queries without a preceding invocation phrase, leading to potential privacy and resource inefficiencies, especially when interacting with secondary assistants of varying trust levels.
Innovation Solution
A primary automated assistant determines a trust level for secondary assistants and processes user queries based on this trust level, selectively providing unmodified, obfuscated, or masked content to ensure security and efficiency in interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the primary automated assistant provides full unmodified audio data to all secondary automated assistants, then the secondary assistants can process queries accurately, but user privacy and data security are compromised for assistants with lower trust levels
Solution Approach 1:
The patent applies local quality by providing different levels of data processing to different secondary automated assistants based on their trust levels. High-trust assistants receive full unmodified audio data for accurate processing, while low-trust assistants receive obfuscated or masked versions with reduced information content. This differential treatment based on local trust characteristics resolves the contradiction between maintaining processing accuracy and protecting user privacy.
Solution Approach 2:
The system changes the parameter of data fidelity based on the trust level of the secondary assistant. By dynamically adjusting the degree of obfuscation or masking applied to audio data according to trust parameters, the system enables high-trust assistants to process queries with full accuracy while limiting low-trust assistants to processed versions, thereby balancing privacy protection with functional effectiveness.
2Object-affected harmful factors
If the primary automated assistant processes and obfuscates audio data for all secondary automated assistants, then user privacy is protected, but the performance and response quality of secondary assistants deteriorate
Solution Approach 1:
The patent implements local quality by tailoring the level of data processing to each secondary assistant's trust level. Rather than uniformly obfuscating all data, the system provides unmodified audio data to high-trust assistants who can deliver high-quality responses, while applying obfuscation only to low-trust assistants. This selective approach maintains response quality for reliable assistants while protecting privacy from less trustworthy ones.
Solution Approach 2:
The system applies partial action by providing full-fidelity audio data only to the extent necessary for high-trust assistants, while applying obfuscation for low-trust assistants where full fidelity is not warranted. This partial provision of unmodified data based on trust levels optimizes the balance between privacy protection and maintaining adequate response quality.
3Speed
If the primary automated assistant provides full audio data to secondary assistants without trust verification, then processing speed is maintained, but security and resource efficiency are reduced
Solution Approach 1:
The patent applies preliminary action by determining the trust level of secondary automated assistants before providing audio data. This pre-assessment of trustworthiness allows the system to make informed decisions about data provisioning, ensuring that high-trust assistants receive full data for fast processing while low-trust assistants receive obfuscated versions, thereby maintaining security without significantly impacting processing speed for reliable assistants.
Solution Approach 2:
The system changes the parameter of data fidelity based on trust level assessments. By dynamically adjusting whether to provide unmodified or obfuscated audio data according to the trust parameters of secondary assistants, the system optimizes the balance between processing speed and security, allowing fast processing for trusted assistants while protecting data from untrusted ones.
Data Source
AI summary
Systems and methods for obfuscating and/or omitting potentially sensitive information in a spoken query before providing the query to a secondary automated assistant. A general automated assistant may be invoked by a user, followed by a query. The audio data can be processed to omit and/or obfuscate potentially sensitive information before providing one or more processed queries to secondary automated assistants based on a trust metric associated with each of the secondary automated assistants. The trust metric for a secondary automated assistant is indicative of trust in being provided with sensitive information. In response, the automated assistants can generate responses, which can be filtered to provide a response to the user.