Encrypted Content Recommendation System for Privacy Preservation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users have privacy and confidentiality concerns regarding their browsing behavior being tracked by online content sources, which can compromise their sensitive information, and existing solutions fail to adequately protect user privacy while recommending content.
Innovation Solution
A method and system that encrypts user attributes and content items using a key, allowing for secure retrieval of targeted content items without revealing sensitive information to online content sources, ensuring privacy through Oblivious Transfer techniques and Homomorphic encryption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If online content sources track user browsing behavior to recommend content, then content recommendation accuracy is improved, but user privacy and confidentiality are compromised
Solution Approach 1:
The patent introduces encrypted user attributes as an intermediary between the user and the content recommendation system. The system processes encrypted representations of user browsing behavior rather than plain text data, allowing recommendation generation while maintaining user privacy through cryptographic protection throughout the processing pipeline.
Solution Approach 2:
The patent transforms user attributes from their original plaintext form into encrypted form through cryptographic operations. This parameter change from readable data to encrypted data allows the system to work with user information in a way that preserves privacy while still enabling recommendation functionality through pattern recognition on encrypted data.
2Object-affected harmful factors
If user attributes are encrypted to protect privacy, then user confidentiality is improved, but content item retrieval capability deteriorates
Solution Approach 1:
The patent creates encrypted copies of user attributes that retain the essential information structure needed for content matching while being cryptographically protected. These encrypted copies enable the system to perform retrieval operations without accessing the original plaintext user data, thus maintaining both privacy and functionality.
Solution Approach 2:
The patent applies cryptographic transformations to user attributes while preserving the semantic relationships needed for content recommendation. The encryption process changes the representation of data parameters but maintains the underlying patterns that enable effective content retrieval and matching.
3Reliability
If content items are encrypted using user keys, then security is improved, but system complexity increases
Solution Approach 1:
The patent implements a system where users manage their own encryption keys and control their data protection. This self-service approach to key management reduces the need for complex centralized security infrastructure, as each user independently secures their own attributes, simplifying the overall system architecture while maintaining high security standards.
Data Source
AI summary
The disclosed embodiments illustrate methods and systems for identifying a targeted content item for a user. The method includes receiving one or more encrypted first attributes of the user, and a first key. Thereafter, one or more content items are encrypted using the first key. The one or more content items are stored in a data structure such that the one or more content items are indexed in the data structure according to one or more second attributes of the one or more content items. Thereafter, at least one encrypted content item is retrieved from the data structure based on the one or more encrypted content items, the indexing of the one or more content items, and the one or more encrypted first attributes. The at least one encrypted content item is decrypted to generate the targeted content item.