Efficiently Computing Homomorphic Matches of Hybrid Pattern Queries on Large Graphs
Document Type
Conference Proceeding
Publication Date
1-1-2019
Abstract
In this paper, we address the problem of efficiently finding homomorphic matches for hybrid patterns over large data graphs. Finding matches for patterns in data graphs is of fundamental importance for graph analytics. In hybrid patterns, each edge may correspond either to an edge or a path in the data graph, thus allowing for higher expressiveness and flexibility in query formulation. We introduce the concept of answer graph to compactly represent the query results and exploit computation sharing. We design a holistic bottom-up algorithm called GPM, which greatly reduces the number of intermediate results, leading to significant performance gains. GPM directly processes child constraints in the given query instead of resorting to a post-processing procedure. An extensive experimental evaluation using both real and synthetic datasets shows that our methods evaluate hybrid patterns up to several orders of magnitude faster than existing algorithms and exhibit much better scalability.
Identifier
85077117619 (Scopus)
ISBN
[9783030275198]
Publication Title
Lecture Notes in Computer Science Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics
External Full Text Location
https://doi.org/10.1007/978-3-030-27520-4_20
e-ISSN
16113349
ISSN
03029743
First Page
279
Last Page
295
Volume
11708 LNCS
Grant
61872276
Fund Ref
National Natural Science Foundation of China
Recommended Citation
Wu, Xiaoying; Theodoratos, Dimitri; Skoutas, Dimitrios; and Lan, Michael, "Efficiently Computing Homomorphic Matches of Hybrid Pattern Queries on Large Graphs" (2019). Faculty Publications. 8000.
https://digitalcommons.njit.edu/fac_pubs/8000
