Search Indexing via Superordinate Axis Aggregation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing search techniques face difficulties in efficiently searching for word or character strings indicated by a search condition at high speeds, as they require verifying the consecutive positions of each word or character, leading to slower search processing.
Innovation Solution
The method involves generating second index information with a superordinate axis based on first index information, aggregating bit strings for each item or register capacity, to narrow down the search area and improve search speed for word or character strings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing search techniques verify consecutive positions of each word or character, then search accuracy is maintained, but search processing speed deteriorates
Solution Approach 1:
The patent segments the search process into two distinct phases: first, rapidly identifying candidate documents containing all required words using bitmap indexes without position verification; second, verifying consecutive positions only within the narrowed-down candidate set. This segmentation allows the majority of documents to be filtered quickly without position checks, while maintaining accuracy through targeted verification on a small subset of candidates.
Solution Approach 2:
The patent performs preliminary filtering using bitmap indexes to identify candidate documents containing all search words before conducting position verification. The bitmap index pre-computes word presence information, allowing the system to quickly eliminate documents that don't contain all required words without performing expensive position verification operations on every document.
2Measurement precision
If existing search techniques process all documents for position verification, then complete accuracy is achieved, but processing time increases
Solution Approach 1:
The patent divides the document set into two groups: candidate documents identified by bitmap index matching, and all other documents. Position verification is applied only to the candidate group, which is a small subset of all documents. This segmentation eliminates unnecessary verification operations on documents that clearly don't contain all search words, significantly reducing total processing time while maintaining accuracy for the candidate set.
Solution Approach 2:
The patent applies position verification partially - only to the extent necessary for documents that pass the bitmap index filter. Rather than verifying positions in all documents (excessive action), the system performs verification only on the minimal necessary subset of candidate documents, achieving complete accuracy for the final result while minimizing processing time.
3Adaptability or versatility
If bitmap tables are obtained from uppermost to lowermost stage in sequence, then multi-stage layered index search is performed, but search efficiency deteriorates due to sequential processing
Solution Approach 1:
The patent merges the bitmap index lookup operations for multiple search words into a single parallel processing step. Instead of sequentially obtaining and processing bitmap tables for each word in the search query, the system retrieves and combines all necessary bitmap data in parallel, then performs the filtering operation simultaneously. This merging eliminates the sequential bottleneck while maintaining the multi-stage index structure's versatility.
Data Source
AI summary
A search method includes receiving a search request to encoded text data, based on first index information produced by specifying an occurrence position of a character or a word included in original data of the encoded text data as a first axis and contents of the search request, generating second index information having a second axis superordinate to the first axis, and searching the encoded text data in response to the search request using the second index information.