Greedy Search Approach of Graph Mining
Greedy search is an efficient and effective strategy for searching an intractably large space when sufficiently informed heuristics are available to guide the search. The space of all subgraphs of a graph is such a space. Therefore, the greedy search approach of graph mining uses heuristics to focus the search toward subgraphs of interest while avoiding search in less interesting portions of the space. One such heuristic is based on the compression afforded by a subgraph; that is, how much is the graph compressed if each instance of the subgraph is replaced by a single vertex. Not only does compression focus the search, but it has also been found to prefer subgraphs of interest in a variety of domains.
Motivation and Background
Many data mining and machine learning methods focus on the attributes of entities in the domain, but the relationships between these entities also represents a significant source of information, and ultimately, knowledge. Mining this relational...
- Cook, D., & Holder, L. (March/April 2000). Graph-based data mining. IEEE Intelligent Systems, 15(2), 32–41.Google Scholar
- Cook, D., Holder, L., Su, S., Maglothin, R., & Jonyer, I. (July/August 2001). Structural mining of molecular biology data. IEEEEngineering in Medicine and Biology, Special Issue on Genomics and Bioinformatics, 20(4), 67–74.Google Scholar
- Eberle, W., & Holder, L. (2006). Detecting anomalies in cargo shipments using graph properties. In Proceedings of the IEEE intelligence and security informatics conference, San Diego, CA, May 2006.Google Scholar
- Gonzalez, J., Holder, L., & Cook D. (2002). Graph-based relational concept learning. In: Proceedings of the nineteenth international conference on machine learning, Sydney, Australia, July 2002.Google Scholar
- Kukluk, J., Holder, L., & Cook, D. (2007). Inference of node replacement graph grammars. Intelligent Data Analysis, 11(4), 377–400.Google Scholar
- Kuramochi, M., & Karypis, G. (2001). Frequent subgraph discovery. In Proceedings of the IEEE international conference on data mining (ICDM) (pp. 313–320), San Jose, CA.Google Scholar
- Matsuda, T., Motoda, H., Yoshida, T., & Washio, T. (2002). Mining patterns from structured data by beam-wise graph-based induction. In Proceedings of the fifth international conference on discovery science (pp. 323–338), Lubeck, Germany.Google Scholar
- Nijssen, S., & Kok, J. N. (2004). A quickstart in frequent structure mining can make a difference. In Proceedings of the tenth ACM SIGKDD international conference on knowledge discovery and data mining (KDD) (pp. 647–652), Seattle, WA.Google Scholar
- Yan, X., & Han, J. (2002). gSpan: Graph-based substructure pattern mining. In Proceedings of the IEEE international conference on data mining (ICDM) (pp. 721–724), Maebashi City, Japan.Google Scholar
- You, C., Holder, L., & Cook, D. (2006). Application of graph-based data mining to metabolic pathways. In Workshop on data mining in bioinformatics, IEEE international conference on data mining, Hong Kong, China, December 2006.Google Scholar