Fuzzy Keyword Search Over Encrypted Data Using Filter Tree
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing searchable encryption systems fail to efficiently perform fuzzy searches in encrypted data due to high overhead, false positives, and fixed fuzziness levels, making them incompatible with dynamic search queries and plaintext domain requirements.
Innovation Solution
A method utilizing a filter tree data structure and function-hiding inner product encryption to enable secure and efficient fuzzy symmetric searchable encryption, allowing for dynamic fuzziness levels and privacy-preserving comparisons without revealing plaintext contents to the server.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If wildcard-based fuzzy sets are used to enable fuzzy search in encrypted data, then fuzzy search capability is achieved, but storage overhead for the search index increases significantly
Solution Approach 1:
The patent segments the fuzzy search index into multiple components: an encrypted exact match index for efficient storage, and a separate encrypted fuzzy search index using LSH functions. This segmentation allows the system to maintain compact storage for exact matches while providing fuzzy search capability through a different index structure that only stores necessary fuzzy search data.
Solution Approach 2:
The patent introduces an intermediary LSH (Locality Sensitive Hashing) function layer between the encrypted data and the fuzzy search query. This intermediary transforms the fuzzy search problem into an exact match problem in the encrypted domain, allowing the server to perform fuzzy searches without storing all possible fuzzy variations, thus reducing storage overhead while maintaining adaptability.
2Productivity
If LSH based methods are used for fuzzy keyword search in encrypted data, then search efficiency is improved, but false positive and negative rates increase in search results
Solution Approach 1:
The patent implements a feedback mechanism where the server performs initial fuzzy search using LSH functions to get candidate results, then returns these candidates to the client for verification. The client uses their decryption key to verify the actual similarity and filter out false positives, creating a feedback loop that improves reliability while maintaining server-side efficiency.
Solution Approach 2:
The patent applies partial action by using LSH functions to generate a relaxed set of candidate results that may include false positives, then performs a more precise verification step only on these candidates. This approach accepts some initial imprecision to gain search efficiency, then corrects the errors through a targeted verification process rather than requiring perfect precision throughout the entire search process.
3Ease of manufacture
If predefined similarity metric is built into the search index structure, then encrypted fuzzy search is enabled, but the system cannot accommodate dynamic fuzziness levels in individual queries
Solution Approach 1:
The patent introduces dynamics by separating the fixed index structure from the query-time fuzziness parameter. The search index maintains a static encrypted structure with LSH functions, while individual queries can specify different fuzziness levels dynamically. The system adjusts the search behavior based on query parameters rather than being fixed by the index structure, allowing flexible adaptability.
Solution Approach 2:
The patent applies parameter changes by allowing the fuzziness level to be varied as a query parameter independent of the index structure. Different queries can specify different edit distance thresholds or similarity criteria, and the system adapts its search behavior accordingly while using the same encrypted index structure. This separates the structural components from the operational parameters, enabling both ease of manufacture and dynamic adaptability.
4Ease of operation
If the server performs fuzzy search operations on encrypted data, then search functionality is provided, but the server cannot identify the content of search terms
Solution Approach 1:
The patent uses encrypted LSH functions as an intermediary that allows the server to perform fuzzy search operations on encrypted data without decrypting it. The server operates on the encrypted representations of data and queries, maintaining privacy while enabling search functionality. The intermediary layer permits computational operations on encrypted data without revealing the underlying plaintext contents to the server.
Data Source
AI summary
A method for performing a fuzzy search in encrypted data includes receiving an encrypted search token corresponding to a search keyword with an untrusted server computing device and generating inner product values based on a function-hiding inner product encryption operation of the encrypted query vectors encrypted node vectors in an encrypted tree stored in the memory of the untrusted server computing device. The method further includes transmitting, with the untrusted server computing device, the encrypted keyword stored in the leaf node to a client computing device in response to the first inner product value exceeding a first predetermined similarity threshold corresponding to a similarity of the first query vector to the leaf node vector, the fuzzy search not revealing plaintext contents for any of a keyword stored in the leaf node, the search keyword, or a fuzziness parameter.