Subgraph query is an important problem in the research of graph databases

and many methods about subgraph query are based on filtering-verification strategy

which key target is to find effective feature patterns.Through the analysis of the embedding information of feature patterns in the data graphs

we propose to construct embedding relation indexing in the offline stage

and propose a new feature pattern embedding based subgraph query algorithm ERSearch.When query graph is given

we will use the co-occurrence relations and embedding relations combined to prune the unmatched data graphs

and the comparing results of embedded relationship in filtering phase can be used in the verification process

improving the efficiency of the verification.Via the experiment in the real and synthetic datasets

compared with PathIndex and other methods

we show that our algorithm can effectively reduce the size of candidate set

and effectively improve the efficiency of filtering and verification stages.