|
| | | |
Continuously Maintaining Order Statistics over Data Streams
Lin, X.
A rank query is essentially to find a data element with a given rank against a monotonic order specified on data elements. It has several equivalent variations [8, 17, 30]. Rank queries over data streams have been investigated in the form of quantile computation. A _-quantile (_ 2 (0, 1]) of a collection of N data elements is the element with rank d_Ne against a monotonic order specified on data elements. Rank and quantile queries have many applications [1, 3, 6, 7, 10, 14-16, 26, 27], including monitoring high speed networks, trends and fleeting opportunities detection in the stock market, sensor data analysis,Web ranking aggregation and log mining, etc. In these applications, they not only play very important roles in the decision making but also have been used in summarizing data distributions of data streams. |
Cite as: Lin, X. (2007). Continuously Maintaining Order Statistics over Data Streams. In Proc. Eighteenth Australasian Database Conference (ADC 2007), Ballarat, Australia. CRPIT, 63. Bailey, J. and Fekete, A., Eds. ACS. 7-10. |
(from crpit.com)
(local if available)
|
|