Data streams and data synopses for massive data sets

Abstract

With the proliferation of data intensive applications, it has become necessary to develop new techniques to handle massive data sets. Traditional algorithmic techniques and data structures are not always suitable to handle the amount of data that is required and the fact that the data often streams by and cannot be accessed again. A field of research established over the past decade is that of handling massive data sets using data synopses, and developing algorithmic techniques for data stream models. We will discuss some of the research work that has been done in the field, and provide a decades' perspective to data synopses and data streams. © Springer-Verlag Berlin Heidelberg 2005.