Access Control Insight Pruning for Privacy

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current information retrieval and productivity systems face challenges in generating insights about relevant content when data is under access control, as changes in access permissions can impact the insights presented to users, potentially leaking private information or tantalizing users with inaccessible content.

Innovation Solution

A computer-implemented method that pre-generates insights for user-content combinations and prunes them based on access control policies at query time, ensuring only accessible related content is suggested to users, thereby respecting changing access controls and maintaining user privacy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If insights are pre-generated for user-content combinations, then insight generation efficiency is improved, but access control compliance deteriorates as users may receive insights about content they no longer have access to

Engineering Contradiction:

Improveinsight generation efficiencyVSAvoidaccess control compliance

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system pre-generates insights for user-content combinations in advance, storing them for rapid retrieval. This preliminary action improves efficiency by avoiding real-time generation overhead, while the pre-generated insights are later filtered at query time to ensure access control compliance.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system extracts and removes insights from the pre-generated set that reference content the user no longer has access to. This extraction process separates compliant insights from non-compliant ones, ensuring that only accessible content is presented to the user while maintaining the efficiency benefits of pre-generation.

Inventive Principle:

Principle #2Taking out (Extraction)

2Loss of information

If access control policies are strictly enforced at insight generation time, then user privacy is protected, but insight relevance deteriorates as access permissions may change after insight generation

Engineering Contradiction:

Improveuser privacy protectionVSAvoidinsight relevance

Core Design Contradiction:

Loss of informationVSAdaptability or versatility

Solution Approach 1:

The system dynamically filters insights at query time based on the user's current access permissions rather than using static access control checks at generation time. This dynamic approach ensures privacy protection while adapting to changes in access permissions, maintaining insight relevance even as user rights evolve.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system uses feedback from the user's current access control state to filter and refine the pre-generated insights. By checking current permissions at query time and removing inaccessible content references, the system ensures both privacy compliance and relevance to the user's current access level.

Inventive Principle:

Principle #23Feedback

3Quantity of substance

If all pre-generated insights are presented to the user, then completeness of suggestions is improved, but security deteriorates as inaccessible content may be revealed

Engineering Contradiction:

Improvenumber of content suggestionsVSAvoidinformation leakage

Core Design Contradiction:

Quantity of substanceVSObject-affected harmful factors

Solution Approach 1:

The system extracts and removes insights that reference inaccessible content from the complete set of pre-generated insights. This extraction ensures that only insights about accessible content are presented to the user, preventing information leakage while maintaining a comprehensive set of relevant suggestions.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The system converts the potential harm of presenting inaccessible content into a benefit by using the access control check as an opportunity to enhance security. The filtering process not only prevents information leakage but also provides a clear audit trail of what content the user can and cannot access, improving overall system security.

Inventive Principle:

Principle #22Blessing in disguise (Convert harm into benefit)

Data Source

PatentUS10944756B2Access control

Publication Date: 2021.03.09 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

A method comprising: pre-generating insights for each of a plurality of user-content combinations, each user-content combination comprising a different respective combination one of a plurality of first users and one of a plurality of first pieces of content, wherein each insight specifies a relationship type and other content having that relationship with the respective first content; subsequently receiving a query seeking insights on a target one of the first users and first pieces of content; based on the query, identifying the respective set of insights for that user-content combination; subsequently pruning away one or more insights which specify no related pieces of content to which the target user is permitted access; and outputting at least one of the remaining subset of insights to the target user.