Exploiting duality in summarization with deterministic guarantees

Karras, P; Mamoulis, N; Sacharidis, D

dc.contributor.author	Karras, P	en
dc.contributor.author	Mamoulis, N	en
dc.contributor.author	Sacharidis, D	en
dc.date.accessioned	2014-03-01T02:51:05Z
dc.date.available	2014-03-01T02:51:05Z
dc.date.issued	2007	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/35358
dc.subject	Efficiency	en
dc.subject	Histograms	en
dc.subject	Synopses	en
dc.subject	Wavelets	en
dc.subject.other	Abstracting	en
dc.subject.other	Computational complexity	en
dc.subject.other	Hierarchical systems	en
dc.subject.other	Information systems	en
dc.subject.other	Wavelet analysis	en
dc.subject.other	Histogram construction algorithms	en
dc.subject.other	Histograms	en
dc.subject.other	Space efficiency	en
dc.subject.other	Synopses	en
dc.subject.other	Data mining	en
dc.title	Exploiting duality in summarization with deterministic guarantees	en
heal.type	conferenceItem	en
heal.identifier.primary	10.1145/1281192.1281235	en
heal.identifier.secondary	http://dx.doi.org/10.1145/1281192.1281235	en
heal.publicationDate	2007	en
heal.abstract	Summarization is an important task in data mining. A major challenge over the past years has been the efficient construction of fixed-space synopses that provide a deterministic quality guarantee, often expressed in terms of a maximum-error metric. Histograms and several hierarchical techniques have been proposed for this problem. However, their time and/or space complexities remain impractically high and depend not only on the data set size n, but also on the space budget B. These handicaps stem from a requirement to tabulate all allocations of synopsis space to different regions of the data. In this paper we develop an alternative methodology that dispels these deficiencies, thanks to a fruitful application of the solution to the dual problem: given a maximum allowed error, determine the minimum-space synopsis that achieves it. Compared to the state-of-the-art, our histogram construction algorithm reduces time complexity by (at least) a Blog2n over logfactor and our hierarchical synopsis algorithm reduces the complexity by (at least) a factor of log2B over log+ logn in time and B(1-log B over log n) in space, where *is the optimal error. These complexity advantages offer both a space-efficiency and a scalability that previous approaches lacked. We verify the benefits of our approach in practice by experimentation. © 2007 ACM.	en
heal.journalName	Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining	en
dc.identifier.doi	10.1145/1281192.1281235	en
dc.identifier.spage	380	en
dc.identifier.epage	389	en