Using a pipeline approach to build data cube for large XML data streams

Abstract

XML has become a widely used standard for data representation, distribution and sharing. The concept of the Sensor Web has led to web generated sensor data in many diverse applications where delivery of the sensed data takes place using the Web. In order to obtain useful knowledge from XML sensor data, data warehouse and OLAP applications aimed at providing support for decision making for operational data must be developed. In this paper, we present a pipeline design based OLAP data cube construction framework designated for real time web generated sensor data, transforming sensor data into XML streams conforming to an underlying data warehouse logical model, which constructs corresponding data cubes. As part of this work, we discuss how our cube construction and acceleration strategy improves the efficiency in managing large volumes of XML data. © Springer-Verlag 2013.