关键词：时间序列； 数据挖掘； 符号化表示； 相似性查找
中图分类号：TP301 文献标志码：A 文章编号：1001-3695(2008)08-2328-04
Novel binary symbolic representation of time series for similarity
SUN Mei-yu1,2， FANG Jian-an1
（1.College of Information Science & Technology, Donghua University, Shanghai 201620, China； 2. Dept. of Computer, Shandong Labour
Union Administrators College, Jinan 250100, China）
Abstract:In spite of there are dozens of techniques for producing different variants of the symbolic representation, there still have no known method to calculate the distance in the symbolic space to provide the lower bounding guarantee. This paper proposed a novel bit level symbolic representation called BSAP. The representation was unique in which it allowed dimensionality reduction and it also granted a lower bound distance measure defined on the symbolic representation. The experiments was performed on synthetic, as well as real data sequences to evaluate the proposed method.
Key words：time series; data mining; symbolic representation; similarity search......