Sixth International Conference on Data Mining (ICDM'06) (2006)
Dec. 18, 2006 to Dec. 22, 2006
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/ICDM.2006.78
Jun Li , Fudan University, China
Haofeng Zhou , Fudan University, China
Wei Wang , Fudan University, China
OLAP is supported by more and more environment as a powerful analysis tool. With the rapid development of mobile and wireless technologies, users wish to enjoy the OLAP service on these devices. However, there are many issues on mobile OLAP against the traditional ones, e.g. the transmission bottleneck, unstable network connection, etc. Moreover, the mobile device owners have raised increasing requirements to customize the service such as transmitting the data on demand or ASAP to support their activities. All these challenges provide new chances for OLAP. In this paper, a new mechanism Gradual Cube is proposed to face such challenges. It can reduce the transmission data size, provide customized transmission strategy and enable users to conduct off-line browsing. We assume the users? precision requirement follows some distribution so that three methods, namely random, optimal and heuristic, are developed to customize the transmission plan. The experiments show that such methods are both effective and efficient.
H. Zhou, J. Li and W. Wang, "Gradual Cube: Customize Profile on Mobile OLAP," Sixth International Conference on Data Mining (ICDM'06)(ICDM), Hong Kong, 2006, pp. 943-947.