Traditional sequence similarity measures have a high time complexity and are therefore not suitable for real-time systems. In this paper, we analyze and discuss properties of sequences as a step toward developing more efficient similarity measures that can approximate the similarity of traditional sequence similarity measures. To explore our findings, we propose a method for encoding sequence information as a vector in order to exploit the advantageous performance of vector similarity measures. This method is based on the assumption that events closer to a point of interest, like the current time, are more important than those further away. Four experiments are performed on both synthetic and real-time data that show both disadvantages and advantages of the method. © 2012 Springer-Verlag.
CITATION STYLE
Gundersen, O. E. (2012). Toward measuring the similarity of complex event sequences in real-time. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7466 LNAI, pp. 107–121). https://doi.org/10.1007/978-3-642-32986-9_10
Mendeley helps you to discover research relevant for your work.