The process of Ubiquitous data mining (UDM) allows data stream mining operations to be conducted on handheld devices with limited resources. Algorithms which take advantage of visualisation can assist users in understanding and interpreting data mining results more quickly. However, there are currently no on-line real-time visualisation tools to complement the UDM algorithms. In this paper we investigate the use of visualisation techniques applied to the cluster change detection domain in a UDM environment. We demonstrate a proof of concept implementation for visualising cluster dynamics and cluster change detection.
展开▼
