In this paper, we present a new algorithm, Separator, for accurate and efficient Hierarchical Heavy Hitter (HHH) detection, an emerging research area of data stream mining. Existing algorithms exploit either bottom-up or topdown processing strategy to solve this problem, whereas we propose a novel combination of these two strategies. Based on this strategy and a devised compact data structure, we implement our algorithm. It is theoretically proved to have tight error bound and small space usage. Comprehensive experiments conducted also verify its accuracy and efficiency. © Springer-Verlag Berlin Heidelberg 2007.
CITATION STYLE
Lin, Y., & Liu, H. (2007). Separator: Sifting Hierarchical Heavy Hitters accurately from data streams. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4632 LNAI, pp. 170–182). Springer Verlag. https://doi.org/10.1007/978-3-540-73871-8_17
Mendeley helps you to discover research relevant for your work.