Contextual Data Delivery With Privacy Filtering
Solution Overview
Problem
Existing technologies face challenges in delivering contextual data to computing devices while ensuring data privacy, particularly in mobile communications, where real-time data delivery is critical but privacy concerns need to be addressed.
Innovation Solution
The solution involves configuring computing devices to manage data privacy settings during communication, such as converting speech to text and applying privacy filters, to determine keywords and retrieve contextual data from a server, ensuring that sensitive information is protected based on privacy attributes of connected devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If contextual data is delivered to computing devices in real-time during mobile communications, then the productivity and responsiveness of the system is improved, but data privacy and security concerns are exacerbated
Solution Approach 1:
The system segments the communication data into private portions (masked portions) and public portions (unmasked portions). The privacy-preserving server processes only the unmasked portions while the masked portions remain encrypted and private to the client device, allowing real-time processing without compromising privacy.
Solution Approach 2:
A privacy-preserving server acts as an intermediary between communicating devices. This server processes contextual data in a privacy-preserving manner by analyzing only unmasked portions of the data, enabling real-time contextual delivery while maintaining data privacy through the mediation of encrypted communications.
2Measurement precision
If speech-to-text conversion is performed on all communication data, then the accuracy of keyword extraction is improved, but the processing time and computational resources required increase
Solution Approach 1:
The system performs speech-to-text conversion and keyword extraction only on unmasked portions of the communication data rather than all data. This partial processing approach maintains sufficient keyword extraction accuracy for the relevant portions while significantly reducing overall processing time and computational resources required.
3Object-affected harmful factors
If privacy filters are applied to communication data, then data privacy is protected, but the quantity of available contextual data for analysis is reduced
Solution Approach 1:
The system applies different processing qualities to different portions of the data. Masked portions maintain their privacy properties without processing, while unmasked portions receive full processing including speech-to-text conversion and keyword extraction. This local quality approach preserves privacy where needed while maintaining sufficient contextual data availability where appropriate.
Data Source
AI summary
A method, device, system, or article of manufacture is provided for improved delivery of contextual data to a computing device while preserving data privacy. In one embodiment, a method comprises sending, from a first computing device, to a second computing device, first communication data; receiving, at the first computing device, from the second computing device, second communication data; in response to determining that a data privacy attribute of the second computing device is the same as a predetermined data privacy attribute, determining to protect the second communication data, including: converting the first communication data to a first set of text; converting the second communication data to a second set of text; and applying a privacy filter to the first set of text to generate a first set of filtered text, wherein the privacy filter removes any text that is associated with the second set of text; determining a first keyword from the first set of filtered text; sending, from the first computing device, to a computer, the first keyword; and receiving, at the first computing device, from the computer, contextual data associated with the first keyword.