Mining Graph Evolution Rules

9 years 2 months ago
Mining Graph Evolution Rules
In this paper we introduce graph-evolution rules, a novel type of frequency-based pattern that describe the evolution of large networks over time, at a local level. Given a sequence of snapshots of an evolving graph, we aim at discovering rules describing the local changes occurring in it. Adopting a definition of support based on minimum image we study the problem of extracting patterns whose frequency is larger than a minimum support threshold. Then, similar to the classical association rules framework, we derive graph-evolution rules from frequent patterns that satisfy a given minimum confidence constraint. We discuss merits and limits of alternative definitions of support and confidence, justifying the chosen framework. To evaluate our approach we devise GERM (Graph Evolution Rule Miner), an algorithm to mine all graph-evolution rules whose support and confidence are greater than given thresholds. The algorithm is applied to analyze four large real-world networks (i.e., two so...
Michele Berlingerio, Francesco Bonchi, Björn
Added 27 May 2010
Updated 27 May 2010
Type Conference
Year 2009
Where PKDD
Authors Michele Berlingerio, Francesco Bonchi, Björn Bringmann, Aristides Gionis
Comments (0)