| Autor: |
Yamamoto, Tsubasa, Ozaki, Tomonobu, Ohkawa, Takenao |
| Zdroj: |
Mining Complex Data; 2008, p143-156, 14p |
| Abstrakt: |
In some real world applications, the data can be represented naturally in a special kind of graphs in which each vertex consists of a set of (structured) data such as item sets, sequences and so on. One of the typical examples is metabolic pathways in bioinformatics. Metabolic pathway is represented in a graph structured data in which each vertex corresponds to an enzyme described by a set of various kinds of properties such as amino acid sequence, enzyme number and so on. We call this kind of complex graphs multi-structured graphs. In this paper, we propose an algorithm named FMG for mining frequent patterns in multi-structured graphs. In FMG, while the external structure will be expanded by the same mechanism of conventional graph miners, the internal structure will be enumerated by the algorithms suitable for its structure. In addition, FMG employs novel pruning techniques to exclude uninteresting patterns. The preliminary experimental results with real datasets show the effectiveness of the proposed algorithm. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
