Retaining all the path information for graph reachability queries based on regular expressions

It is common to find that many real world data graphs are edge-labeled, i.e., each edge attaches a label to indicate the relationship of the two vertices connected by the edge. A basic research issue on these graphs is how to make a query to test if the vertex u can reach the vertex v through the path constrained by a series of labels. In this paper, we propose a novel index framework to retain all the path label information of a data graph and answer the regular expression based reachability queries efficiently. Our index is a generalized-list-structured index and constructed based on the method of bread first search and depth first search. To improve the query efficiency, we also put forward a labeling mechanism for our index. A detailed experimental evaluation on both real and synthetic datasets shows that the performance of our method is improved in term of both index size and query time.