SIMILAR TIME-SERIES DATA COMPUTING DEVICE, SIMILAR TIME-SERIES DATA COMPUTING METHOD, AND SIMILAR TIME-SERIES DATA COMPUTING PROGRAM

摘要

PPROBLEM TO BE SOLVED: To keep a computing cost low in determining the similarity of time-series data. PSOLUTION: The time-series data are narrowed down by a narrowing down part 3 by using multi-dimensional index structure, and a distance between dimensionally compressed time-series data is found by a similarity determination part 4, and the data are determined to be similar in the case that the distance is at most a prescribed value. Thus, a computing cost is kept low by conducting similarity determination using the dimensionally compressed time-series data. High-precision dimensional compression is realized by computing the average and standard deviation of the time-series data for each frame as a dimensionally compressing means. PCOPYRIGHT: (C)2007,JPO&INPIT