Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
PROCESSING AGGREGATE QUERIES OVER CONTINUOUS DATA STREAMS 251 ABSTRACT OF PRESENTATION Processing Aggregate Queries over Continuous Data Streams Johannes Gehrke, Cornell University In this talk, I will describe techniques for giving approximate answers for aggregate queries over data streams using probabilistic âsketchesâ of the data streams that give approximate query answers with provable error guarantees. I will introduce sketches and then talk about two recent technical advances, sketch partitioning and sketch sharing. In sketch partitioning, existing statistical information about the stream is used to significantly decrease error bounds. Sketch sharing allows one to improve the overall space utilization among multiple queries. I will conclude with some open research problems and challenges in data stream processing. Part of this talk describes joint work with Al Demers, Alin Dobra, and Mirek Riedewald at Cornell and Minos Garofalakis and Rajeev Rastogi at Lucent Bell Labs.