A Direct Mining Approach To Efficient Constrained Graph Pattern Discovery

• المؤلفون الرئيسيون: ZHU, Feida, ZHANG, Zequn, QU, Qiang
• التنسيق: text
• اللغة: English
• منشور في: Institutional Knowledge at Singapore Management University 2013
• الموضوعات: Direct mining; frequent graph pattern mining; constrained patternmining; skinny pattern; Databases and Information Systems; Numerical Analysis and Scientific Computing
• الوصول للمادة أونلاين: https://ink.library.smu.edu.sg/sis_research/1819; https://ink.library.smu.edu.sg/context/sis_research/article/2818/viewcontent/C53___A_Direct_Mining_Approach_To_Efficient_Constrained_Graph_Pattern_Discovery__SIGMOD2013_.pdf
• المؤسسة: Singapore Management University

Despite the wealth of research on frequent graph pattern mining, how to efficiently mine the complete set of those with constraints still poses a huge challenge to the existing algorithms mainly due to the inherent bottleneck in the mining paradigm. In essence, mining requests with explicitly-specified constraints cannot be handled in a way that is direct and precise. In this paper, we propose a direct mining framework to solve the problem and illustrate our ideas in the context of a particular type of constrained frequent patterns — the “skinny” patterns, which are graph patterns with a long backbone from which short twigs branch out. These patterns, which we formally define as l-long d-skinny patterns, are able to reveal insightful spatial and temporal trajectory patterns in mobile data mining, information diffusion, adoption propagation, and many others. Based on the key concept of a canonical diameter, we develop SkinnyMine, an efficient algorithm to mine all the l-long d-skinny patterns guaranteeing both the completeness of our mining result as well as the unique generation of each target pattern. We also present a general direct mining framework together with two properties of reducibility and continuity for qualified constraints. Our experiments on both synthetic and real data demonstrate the effectiveness and scalability of our approach.