Graph Search Optimization Using Sorted Property Techniques
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large graphs representing complex data relationships are computationally expensive to search for patterns, making it difficult to extract valuable information due to the need to check every vertex for every vertex in the pattern, especially when dealing with billions of entities and relationships, which can result in lengthy processing times that render the extracted information useless.
Innovation Solution
The Graph Search Optimization System (GSOS) employs techniques such as edge-count directed systems, derived constraint systems, and sorted property systems to reduce computational expense by factoring in edge counts, deriving single-element constraints from multi-element constraints, and sorting edges to efficiently search for patterns within graphs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If every vertex of the graph is checked for every vertex of the pattern, then complete pattern matching is achieved, but computational expense increases significantly
Solution Approach 1:
The patent applies preliminary action by pre-computing and storing out-degree and in-degree counts for each vertex before the pattern matching process. This preliminary preparation allows the algorithm to quickly filter candidate vertices during pattern matching without performing exhaustive checks, thereby reducing computational expense while maintaining matching completeness.
Solution Approach 2:
The patent segments the pattern matching process into distinct phases: initializing vertex counts, matching pattern vertices to graph vertices using degree constraints, and verifying edges. This segmentation allows each phase to operate efficiently with optimized data structures and algorithms, reducing overall computational complexity compared to a monolithic approach.
2Measurement precision
If exhaustive graph searching is performed, then all matching patterns are identified, but processing time becomes excessively long
Solution Approach 1:
The patent performs preliminary action by pre-calculating and storing the out-degree and in-degree of each vertex in the graph. During pattern matching, these pre-computed values are used to quickly identify candidate vertices that satisfy degree constraints, avoiding the need to count edges during the matching process itself and significantly reducing processing time.
Solution Approach 2:
The patent applies partial action by using degree constraints to filter and limit the search space to only those vertices that could potentially match pattern vertices. Instead of checking every vertex in the graph, the algorithm focuses computation on a subset of candidate vertices identified by matching degree properties, thereby reducing processing time while ensuring all valid matches are found.
3Productivity
If the graph search is optimized using edge counts and constraints, then processing speed increases, but system complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-computing vertex degree counts and storing them in data structures before the pattern matching process begins. This preliminary preparation enables the matching algorithm to quickly compare pattern vertex degrees with graph vertex degrees without performing expensive edge counting operations during matching, thereby increasing search speed while keeping the added complexity manageable through efficient data storage.
Solution Approach 2:
The patent changes parameters by using vertex degree (out-degree and in-degree) as filtering criteria to constrain the search space. By transforming the problem from checking all vertex pairs to only checking vertices that satisfy degree constraints, the system achieves faster processing. The complexity increase is offset by the efficiency gains from parameter-based filtering and optimized data access patterns.
Data Source
AI summary
Various systems are provided for optimizing the searching of a graph for a portion that matches a pattern is provided. A Graph Search Optimization System (“GSOS”) provides various techniques for reducing the computational expense when searching for patterns within a graph. The GSOS provides techniques that include an edge-count directed (“ECD”) system, a derived constraint (“DC”) system, and a sorted property (“SP”) system. The ECD system matches a pattern in a direction based on the number of edges for that direction. The DC system derives a single-element constraint from a multi-element constraints to avoid having to check multiple elements. The SP system processes edges of a graph in a sorted order based on the value of a property of the edges.