Graph Search Indexing via Path Fields and Posting Lists
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing information search devices face inefficiencies in searching for target nodes from graph structure information as the complexity increases, leading to prolonged search times due to increased tuple counts in tables and numerous join operations.
Innovation Solution
An information search device that generates path fields and posting lists for each node, associating them in an index repository, allowing for efficient searching by matching sequence appearances in the search path, thereby reducing the number of comparisons and resource consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If graph structure information is stored in traditional table formats with edges as indexes, then complete information can be stored, but the number of tuples increases and search time prolongs as complexity increases
Solution Approach 1:
The patent segments graph structure information into two distinct storage components: a path table storing only path expressions (columns of element values from origin to destination) and an edge table storing only edge index information. This segmentation allows the path table to remain compact while the edge table provides efficient indexing, resolving the contradiction between complete information storage and fast search performance.
2Loss of information
If path tables include all possible path information, then complete path data is available, but the number of tuples increases leading to more comparisons and higher resource consumption
Solution Approach 1:
The patent extracts edge index information from the path table and stores it separately in an edge table. The path table retains only the essential path expressions needed for search matching, while edge information is maintained in a separate indexed structure. This extraction reduces the number of tuples in the path table while preserving complete path information through the combined structure of both tables.
Data Source
AI summary
An information search device includes: a path field generating part that, for each node, generates a path field in which paths with the node as an origin are connected; an index generating part that, for each element configuring graph structure information, generates a posting list as a list of information composed of a node having a path field including the element and location information representing a location where the element appears in the path field, and generates an index repository in which the element and the posting list are associated; a search path generating part that generates a search path representing a search condition; and a searching part that extracts a node having a path field satisfying an appearance sequence of elements included in the search path based on the location information of the posting list.