This paper presents the FCP-Tree index structure and the new algorithm for continuous pattern mining, called FCPGrowth, for Trajectory Data Warehouses. The FCP-Tree is an aggregate tree which allows storing similar sequences in the same nodes. A characteristic feature of the FCPGrowth algorithm is that it does not require constructing intermediate trees at recursion levels and therefore, it has small memory requirements. In addition, when the initial FCP-Tree is built, input sequences are split on infrequent elements, thereby increasing the compactness of this structure. The FCPGrowth algorithm is much more efficient than our previous algorithm, which is confirmed experimentally in this paper. © 2010 Springer-Verlag.
CITATION STYLE
Gorawski, M., & Jureczek, P. (2010). Continuous pattern mining using the FCPGrowth algorithm in trajectory data warehouses. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 6076 LNAI, pp. 187–195). https://doi.org/10.1007/978-3-642-13769-3_23
Mendeley helps you to discover research relevant for your work.