Graph-structured data compression based on frequent subgraph contraction

Abstract

There is much redundant information in graph-structured which has self-describing characteristic, so how to compress graph-structured so as to improve the efficiency of management of data is becoming more and more significant. This paper studies storage oriented graph-structured compression techniques. For the graph given, many subgraph will be generated. A based on graph traversal, the frequently patterns(fp) can be found. With join the fp patterns, a new fp pattern is produced. Followed by this loop until the threshold, the it result in compression results. Analysis and experiments show that the algorithms have high performance. © 2012 Springer-Verlag.