Efficiently Processing Multiple Subgraph Matching Queries In Graph Database

Graph database is a novel form of storage which is based on graph theory, de-scribes and stores different kinds of data by using nodes, edges and their attributes. Graph database has special advantages on large, complex, semi-structured or un-structured data, and is a hot research topic in database field. As a basic problem in graph database, subgraph matching has drew scholars’ attention, but the ex-isting algorithms are focus on optimizing the single query. With widespread use of the graph database in the field of social networks, and biological information, multi-query processing becomes an important research in graph database.This paper focuses on processing multiple subgraph matching queries. There are two kinds of data sources:one is graph database, the other is query stream, so we propose a dual index that database index is built on frequent fragments, and query index is built on fragments in query stream, bind two indices by fragments inclusion relationship, and store some useful information(queries, matching sets and so on) in it. We try to use intermediate results in query history to improve the processing which queries are similar. By using the methods of graph encoding and hash mapping, the dual index can support fast random access, and quick maintenance, and get matching set quickly by bi-direction search. At last, we do some experiments to compare with existing algorithms, verify and analyze that the approach is effective than others.