Relationship Graph Evaluation via Virtual Network Simulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Evaluating the relationship of a node with respect to a reference point node in a relationship graph is inefficient due to the presence of multiple paths and the difficulty in determining the shortest path, especially in graphs with a large number of nodes and links.
Innovation Solution
A relationship graph evaluation system that creates a virtual communication network based on the relationship graph, simulates the propagation of communication packets, and evaluates the relationship based on attributes such as propagation success rate, delay, and reliability, allowing for accurate evaluation of node relationships.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all paths are checked to determine the shortest path, then the relationship evaluation becomes comprehensive, but the computation time and complexity increase significantly
Solution Approach 1:
The patent extracts only the necessary path information (shortest path and second shortest path) rather than checking all possible paths. This selective extraction maintains evaluation accuracy while dramatically reducing computation time by focusing only on the most relevant paths for relationship assessment.
Solution Approach 2:
The patent performs partial action by checking only a limited number of paths (specifically the shortest and second shortest) instead of all possible paths. This partial checking is sufficient to achieve accurate relationship evaluation without the excessive computational burden of exhaustive path analysis.
2Productivity
If the shortest path is used to determine node relationships, then the evaluation is computationally efficient, but the relationship may not be accurately represented
Solution Approach 1:
The patent goes slightly beyond using only the shortest path by also considering the second shortest path. This partial extension maintains computational efficiency while improving relationship representation accuracy, especially for nodes where the shortest path may not fully capture the relationship dynamics.
Solution Approach 2:
The patent changes the evaluation parameter from considering only path length to considering both path length and path diversity (shortest and second shortest paths). This parameter change allows the system to maintain efficiency while achieving more accurate relationship representation.
3Measurement precision
If multiple paths are considered for relationship evaluation, then the evaluation accuracy improves, but the complexity of path checking increases
Solution Approach 1:
The patent extracts only the essential path information (shortest and second shortest paths) needed for accurate relationship evaluation, ignoring all other paths. This selective extraction improves evaluation accuracy while keeping the checking complexity manageable by focusing on a limited set of critical paths.
Data Source
AI summary
The invention aims to provide a relationship evaluation system capable of effectively evaluating a relationship of a node to a reference node in a relationship graph in which plural nodes are connected by links. A reference point node is selected based on selection information regarding the reference point node input from the selection information input unit. Based on the reference point node, a relationship graph becoming an evaluation target is extracted from the relationship graph database to create a virtual communication network based on the graph. In the virtual communication network, simulation is executed by sequentially propagating a virtual communication packet to adjacent nodes from a reference point node selected by the reference point node selection unit as propagation origin. The relationship of the node to the reference node is evaluated based on the content of the communication packet received at the node as a result of executing the simulation.