Frequent subgraph summarization with error control

Abstract

Frequent subgraph mining has been an important research problem in the literature. However, the huge number of discovered frequent subgraphs becomes the bottleneck for exploring and understanding the generated patterns. In this paper, we propose to summarize frequent subgraphs with an independence probabilistic model, with the goal to restore the frequent subgraphs and their frequencies accurately from a compact summarization model. To achieve a good summarization quality, our summarization framework allows users to specify an error tolerance σ, and our algorithms will discover k summarization templates in a top-down fashion and keep the frequency restoration error within σ. Experiments on real graph datasets show that our summarization framework can effectively control the frequency restoration error within 10% with a concise summarization model. © 2013 Springer-Verlag Berlin Heidelberg.