Reducing Search Space in Subgraph Matching Problem

Abstract

Subgraph matching problem refers to finding query graphs in a large graph. The size of search space in subgraph matching depends on the size of large graph. Due to this large search space, some methods have been proposed to reduce the computational time of matching by preprocessing the large graph. The structural indexing methods restrict the potential occurrences of subgraphs. However, a large percent of these candidates are false positives, which waste resources in matching time. In this paper, we propose a method to find and remove false positive candidates using spectral features in localities. Experiments on biological datasets demonstrate the efficiency of our method in terms of pruning the search space and reducing the matching time.