Document Encryption Index for Secure Remote Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional document processing systems face challenges in ensuring information security when storing documents remotely, as they either require unreliable decryption capabilities in the remote server or consume excessive bandwidth by downloading and decrypting large encrypted documents locally.
Innovation Solution
A document processing method where a client divides documents into pages, encrypts them with a first key, extracts significant words, encrypts these words and page information with a second key, and transmits both to a remote server, preventing the server from decrypting the data and enhancing security, while allowing efficient keyword searching and data processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the remote server has decryption capability to perform keyword searching, then data processing demand is satisfied, but information security is compromised as the server can peek/leak documents
Solution Approach 1:
The patent segments the encryption system into two independent parts: document pages encrypted with a first key and significant words/index encrypted with a second key. The remote server only possesses the second key and can only search the encrypted index, not decrypt the actual document content. This segmentation allows the server to perform search operations while maintaining security isolation between the search capability and the document content.
Solution Approach 2:
The patent introduces an encrypted index structure as an intermediary between the client's search request and the document storage. The index contains encrypted significant words and their corresponding page locations, allowing the server to perform keyword matching without decrypting documents. This intermediary enables data processing while preventing direct access to sensitive information.
2Reliability
If the remote server does not have decryption capability, then information security is maintained, but bandwidth resources are consumed excessively as the client must download complete encrypted documents for processing
Solution Approach 1:
The patent extracts only the essential search-related information (significant words and their page locations) from the complete document and stores it in an encrypted index structure. When a search is needed, the client only needs to download this compact index rather than the entire document, dramatically reducing bandwidth consumption while maintaining security.
Solution Approach 2:
The patent performs preliminary processing by pre-extracting and pre-encrypting significant words and their location information into an index structure before storage. This preliminary action creates a searchable representation that is much smaller than the original document, enabling efficient remote processing without transferring large amounts of data.
3Productivity
If the client downloads and decrypts complete encrypted documents locally for processing, then data processing requirements are met, but bandwidth resources are wasted on transmitting large amounts of encrypted data
Solution Approach 1:
The patent creates a simplified encrypted copy (the index) that contains only the essential search information needed for data processing. This copy is much smaller than the original document but sufficient for keyword searching and location retrieval, eliminating the need to transmit and process the complete document data.
Data Source
AI summary
A document processing method and system divides a document into document pages, and encrypts the document pages by first key to obtain a plurality of encrypted pages; picks a part of words from the document pages and encrypts them by second key to obtain a Significant Word Set (SWS); picks a part of words from the picked part of words and encrypts them by third key to obtain a Most Relevant Word Set (MRWS). The encrypted pages, the SWS and the MRWS are transmits to a remote server for storage. When user search a keyword in the document, the keyword is encrypted by the second and third keys for performing two query. The first query result is decrypted to obtain the search result. The second query result is decrypted and then checked whether it is a subset of the first decrypted query result for detecting unfaithful execution.