Local Query Feature Learning for Privacy-Preserving Search Personalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current search engines face challenges in providing user-specific, relevant results while respecting user privacy, as they need to gather extensive user information to personalize queries, which raises privacy concerns.
Innovation Solution
Implementing local learning on a user's computing device to analyze query features and user feedback, allowing the device to train machine learning algorithms using private information without sharing it with remote search engines, thereby enhancing search results relevance while maintaining user privacy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a search engine gathers extensive information about users to personalize queries, then search result relevance is improved, but user privacy is compromised
Solution Approach 1:
The patent segments the search system into local and remote components. Local learning algorithms run on the user's device to extract query features, while remote search engines receive only anonymized feature vectors rather than raw user data. This segmentation allows personalization while preserving privacy by separating data processing from data storage.
Solution Approach 2:
The patent introduces an intermediary layer of local feature extraction and vectorization. Instead of directly sharing raw user queries and search history with remote servers, the system uses local machine learning models to transform personal data into anonymized feature representations that serve as intermediaries between user behavior and remote search processing.
2Object-affected harmful factors
If local learning algorithms are implemented on user devices, then user privacy is protected, but the complexity of the search system increases
Solution Approach 1:
The patent applies preliminary action by pre-training machine learning models locally on user devices to extract query features before data leaves the device. This preliminary processing of data into anonymized features reduces the complexity of remote server operations and minimizes the amount of sensitive data that needs to be transmitted and processed centrally.
Solution Approach 2:
The system enables self-service by allowing local learning algorithms to autonomously process user queries and generate feature vectors without requiring centralized processing of raw user data. The user's device independently performs feature extraction and sends only the essential anonymized representations to remote servers, reducing overall system complexity.
3Measurement precision
If query features are extracted and sent to remote search engines, then personalized search results are achieved, but data transmission requirements increase
Solution Approach 1:
The patent extracts only the essential query features and transforms them into compact feature vectors locally before transmission. By taking out only the necessary anonymized feature representations rather than transmitting complete raw user data, the system achieves personalized search results while minimizing data transmission volume.
Solution Approach 2:
The system applies parameter changes by transforming raw user query data into different parameter representations (feature vectors) that preserve the essential information needed for personalization while occupying significantly less space. This parameter transformation reduces data transmission requirements while maintaining the ability to generate personalized search results.
Data Source
AI summary
Systems and methods are disclosed for improving search results returned to a user from one or more domains, utilizing query features learned locally on the user's device. One or more domains can inform a computing device of one or more features related to a search query upon which to the computing device can apply local learning. A local search system can include a local database, a local search history and feedback history database, and a local learning system to identify features about query terms. The features can be learned from the user's interaction with both local search results and remote search results, without sending the user interaction information or other user identification information to a remote search engine. A locally learned feature can be used to extend a query, bias a query term, or filter query results.