Search Query Segmentation for Index Intersection Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Character index systems face inefficiencies in searching for indexed documents due to the high computational cost of set intersection operations, which are exacerbated by the large volume of indexed documents and multiple search terms in queries.
Innovation Solution
Implementing a method that segments search queries into terms, processes subsets of indexed document sets using selection operations to reduce the number of set intersection operations and documents processed, thereby enhancing search efficiency by selecting the smallest document sets for each term and performing intersections only on these subsets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If set intersection operations are performed on all retrieved indexed document sets for each search term, then complete search results are obtained, but computational complexity and processing time increase significantly
Solution Approach 1:
The patent segments the search query into multiple search terms, and for each search term, retrieves indexed document sets from relevant index tables. Then, instead of performing set intersection on all document sets simultaneously, it processes them in groups, performing intersections only on document sets that share common index tables. This segmentation approach reduces the computational complexity while maintaining search result completeness.
Solution Approach 2:
The patent applies partial action by selecting only the necessary indexed document sets for processing. Instead of performing set intersection on all retrieved document sets, it identifies and processes only those document sets that are needed based on the search terms and index table relationships. This partial processing approach maintains sufficient search results while significantly reducing computational overhead.
2Quantity of substance
If multiple index tables are queried for each search term, then comprehensive indexed documents are retrieved, but the number of set intersection operations increases
Solution Approach 1:
The patent merges the retrieval process by querying multiple index tables for the same search term simultaneously, then combining the retrieved indexed document sets. It identifies common index tables across different search terms and merges the corresponding document sets before performing set intersection operations. This merging approach reduces the total number of set intersection operations while maintaining comprehensive document retrieval.
Solution Approach 2:
The patent performs preliminary action by pre-identifying which index tables are common across multiple search terms before executing the full search process. It analyzes the search terms and determines in advance which indexed document sets can be processed together, thereby preparing the search structure to minimize the number of set intersection operations needed during execution.
3Measurement precision
If all indexed document sets are processed for each search term, then accurate search results are obtained, but processing time increases significantly
Solution Approach 1:
The patent applies local quality by processing indexed document sets differently based on their specific characteristics and relationships. Instead of uniformly processing all document sets, it identifies local patterns such as common index tables and processes document sets with similar properties together. This localized processing approach maintains search accuracy while reducing overall processing time by avoiding redundant operations on document sets that can be handled efficiently as groups.
Data Source
AI summary
Finding indexed documents is described herein, including: extracting a character string from the search query and segmenting the character string into a plurality of search terms, wherein each search term comprises one or more search words; retrieving indexed document sets corresponding to search words included in the plurality of search terms; processing the retrieved indexed document sets associated with a first search term of the plurality of search terms based at least in part on a selection operation, wherein processing results of the retrieved indexed document sets associated with the first search term comprise search results for the first search term; performing set intersection operations on search results corresponding to the plurality of search terms to generate a results document set; and determining and returning a first subset of indexed documents of the results document set that each includes each of the plurality of search terms.