Graph search device, graph search system, and graph search method

A graph search system in which a graph search device for performing a graph search and a user terminal are connected via a communication network, wherein

• the graph search device comprises:
• a processor;
• a memory;
• and a storage unit,
• the storage unit includes an existing information database storing existing information that includes a set of existing process graphs indicating a flow of a process in graph format and a set of existing process tables associated with the set of existing process graphs and storing information regarding the set of existing process graphs in table format, and
• the memory includes processing instructions for causing the processor to function as:
• a query data management unit that acquires, from the user terminal, query data that is a search query for the existing information and that includes a query graph and a query table;
• a structure comparison unit that compares the query graph included in the query data with the set of existing process graphs to calculate a structural similarity between the query graph and the set of existing process graphs, and determine, as a first search candidate, an existing process graph satisfying a predetermined structural similarity threshold value with respect to the query graph, from among the set of existing process graphs;
• a content comparison unit that compares the query table included in the query data with the set of existing process tables to calculate a content similarity between the query graph and the set of existing process tables, and determine, as a second search candidate, an existing process graph associated with an existing process table satisfying a predetermined content similarity threshold value for the query table, from among the set of existing process tables;
• and
• a results management unit that determines search results data for the query data among the first search candidate and the second search candidate based on a predetermined comprehensive evaluation criterion, and transmits the search results data to the user terminal.

A graph search method, comprising:

• constituting an existing information database storing existing information that includes a set of existing process graphs indicating a flow of a process in graph format and a set of existing process tables associated with the set of existing process graphs and storing information regarding the set of existing process graphs in table format;
• acquiring query data that is a search query for the existing information and that includes a query graph and a query table;
• determining a set of paths present in each existing process graph included in the set of existing process graphs by using a predetermined graph analysis method to analyze the set of existing process graphs;
• determining, for each path included in the set of determined paths, first node name information indicating a node name of each node included in the path for each existing process graph by using a predetermined natural language process method;
• determining second node name information indicating a node name of each node included in the query graph by using a predetermined natural language process method;
• comparing, using a predetermined natural language processing method, the first node name information determined for each existing process graph with the second node name information to calculate a structural similarity between the query graph and the set of existing process graphs, and determine, as a first search candidate, an existing process graph satisfying a predetermined structural similarity threshold value for the query graph, from among the set of existing process graphs;
• determining, for each existing process table, first column name information indicating a column name of a column present in each existing process table included in the set of existing process tables, by using a predetermined natural language processing method;
• determining second column name information indicating a column name of a column included in the query table by using a predetermined natural language processing method;
• comparing, using a predetermined natural language processing method, the first column name information determined for each existing process table with the second column name information to calculate a content similarity between the first column name information and the second column name information, and specify, from within each existing process table, a first column set satisfying a predetermined content similarity threshold value for a second column set included in the query table;
• comparing first table information stored in the first column set specified from within each existing process table with second table information stored in a second column set included in the query table to determine, for each existing process table, content similarity between the first table information and the second table information, and determine, as the second search candidate, an existing process graph associated with an existing process table that includes first table information satisfying a predetermined content similarity threshold value with respect to the second table information, from among the set of existing process tables;
• and
• determining search results data for the query data among the first search candidate and the second search candidate based on a predetermined comprehensive evaluation criterion, and outputting the search results data.