International Journal of
Information and Education Technology

Abstract

In the regime of “Big Data”, data compression techniques take crucial part in preparation phase of data analysis. It is challenging because statistical properties and other characteristics need to be preserved while the size of data need to be reduced. In particular, to compress trajectory data, movement status (such as position, direction, and speed etc.) need to be retained. Moreover, for the increasing demand of real-time processing capability, “online” algorithms are becoming more desirable in data analysis. In this paper, we introduce an On-Line Data Compression Algorithms for Trajectories (OLDCAT), which is an elegant, fast algorithm to effectively compress trajectory data to desirable volume. It is able to deal with real-time data, and scalable to adapt to different sensitivity, accuracy, and compression requirements. An evaluation of its parameter settings and a case study are also discussed in this paper.