Visual Exploration of Stream Pattern Changes Using a Data-Driven Framework

摘要: When using visualization techniques to explore data streams, an important task is convey pattern changes. Challenges include: (1) Most analysis tasks require users observe the change over a long time range; (2) The rate of patterns not constant, and most are normally more interested in bigger changes than smaller ones. Although distorting axis as proposed literature can partially solve this problem, these driven by user. This however applicable streaming exploration that near real-time responsiveness. In paper, we propose data-driven framework merge thus condense windows having small or no Only significant shown users. Juxtaposed views discussed for conveying Our experiments show our algorithm preserves information uniform sampling. We also conducted user study confirm help find quickly via non-distorted axis.