Circuit Diagram Search via Graph Similarity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current search methods for circuit diagrams cannot consider connection information, making it difficult for machines to find network circuit diagrams and requiring human visual search.
Innovation Solution
A search system that converts input circuit diagram information into a graph, calculates similarity with registered graph data, and displays the results, using a server-connected database and methods like the Weisfeiler-Lehman kernel for accurate similarity calculation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If common search methods are used for circuit diagrams, then search speed is maintained, but connection information cannot be considered and search accuracy deteriorates
Solution Approach 1:
The patent replaces traditional text-based or image-based search methods with a graph theory-based approach. Circuit diagrams are converted into graph structures where nodes represent circuit elements and edges represent connections. This substitution enables machine-based semantic understanding of connection information while maintaining automated search functionality, resolving the contradiction between search accuracy and system complexity.
Solution Approach 2:
The patent transforms circuit diagram data from traditional formats into graph theoretical parameters (nodes, edges, adjacency matrices). By changing the representation parameters of circuit information, the system can now consider connection relationships in search operations. This parameter transformation enables accurate matching based on connection topology without requiring complex manual analysis.
2Measurement precision
If graph-based search method is implemented, then connection information can be considered and search accuracy improves, but calculation complexity increases
Solution Approach 1:
The patent segments the circuit diagram analysis into distinct graph theoretical operations: conversion to graph structure, similarity calculation using graph algorithms, and result ranking. By dividing the search process into manageable segments based on graph theory principles, the system achieves accurate connection-based search while controlling calculation complexity through efficient graph algorithms.
Solution Approach 2:
The patent creates graph structure copies of circuit diagrams for comparison purposes. Instead of analyzing original complex circuit diagrams directly, the system works with simplified graph representations that capture essential connection information. This copying approach enables efficient similarity calculations while maintaining search accuracy based on connection relationships.
3Loss of information
If visual search by human is used, then connection information can be analyzed, but search time increases and productivity decreases
Solution Approach 1:
The patent enables the search system to automatically analyze connection information without requiring human visual inspection. By implementing graph-based algorithms that inherently understand connection relationships, the system performs self-service analysis of circuit diagram connections. This automation eliminates the need for human experts to manually analyze connection information, thereby maintaining comprehensive analysis capability while dramatically improving search efficiency and productivity.
Data Source
AI summary
A search system and a search method are provided. The search system includes a means to input first circuit diagram information to a terminal, a database in which second circuit diagram information is registered, a means to convert the input first circuit diagram information into a first graph, an arithmetic means to calculate similarity between the first graph and a second graph based on the second circuit diagram information, and a means to display the similarity on the terminal.