Highlight Index Segmentation for Secure Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional search services that store document contents pose security risks for organizations, as they require external search services to have access to sensitive information, necessitating a method to highlight documents without compromising security.
Innovation Solution
Implementing a system where an internal search service creates a highlight index with an analyzer for document types, allowing for search query processing and highlighting without storing complete document contents, thereby separating the highlight service from the external search service.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If an external search service stores complete document contents to enable highlighting, then search result highlighting capability is improved, but security risk increases due to exposure of sensitive information
Solution Approach 1:
The patent segments the search system into two separate services: an external search service that handles only metadata and document identifiers, and an internal search service that stores and processes complete document contents. This segmentation allows the external service to provide highlighting functionality without exposing sensitive document contents, as it only receives processed highlight queries and returns highlight positions without accessing actual document content.
Solution Approach 2:
The patent introduces an intermediary internal search service that acts as a mediator between the external search service and the document storage system. The internal service receives highlight queries from the external service, performs the highlighting operations on stored documents, and returns only the highlight positions and snippets without exposing the full document contents to the external service, thus maintaining security while enabling highlighting.
2Measurement precision
If an external search service accesses complete document contents to provide highlighting, then search result quality is improved, but information exposure increases
Solution Approach 1:
The system segments information access rights by separating the external search service from complete document contents. The external service only accesses metadata, document identifiers, and pre-generated highlight snippets, while the internal service maintains access to complete documents for processing highlight queries, thereby preventing information exposure while maintaining search quality.
Solution Approach 2:
The patent creates a simplified copy of the search system architecture where the external search service operates on a subset of data (metadata and highlight snippets) rather than complete documents. This copying approach allows the external service to provide high-quality search results with highlighting functionality while working only with non-sensitive data copies.
3Manufacturing precision
If document contents are stored in external search service for highlighting, then highlighting accuracy is improved, but security vulnerability increases
Solution Approach 1:
The patent segments the system architecture to place complete document contents exclusively in the internal search service, while the external service operates only with metadata and highlight results. This segmentation ensures that highlighting accuracy is maintained through proper indexing and query processing in the internal service, while the external service never stores or processes sensitive document contents, eliminating the security vulnerability.
Solution Approach 2:
The internal search service acts as an intermediary that processes all highlighting operations on complete documents and returns only sanitized results (highlight positions and snippets) to the external service. This intermediary approach ensures accurate highlighting is performed on complete documents while the external service remains vulnerable-free as it never handles sensitive content.
Data Source
AI summary
The disclosed computer-implemented method for enabling search services to highlight documents may include (1) creating, via an internal search service, a highlight index that comprises an analyzer for at least one type of document, (2) receiving a search query configured for an external search service and a document that is of the type and that comprises a search result for the search query, (3) querying the highlight index in order to retrieve the analyzer for the type of document from the highlight index, and (4) sending the analyzer, the document, and the search query to a search service in order to enable the search service to display at least one highlighted string extracted from the document via the analyzer, wherein the highlighted string originates from the search query. Various other methods, systems, and computer-readable media are also disclosed.