Title: Detection of variable length anomalous subsequences in data streams

Authors: Amany Abou Safia; Zaher Al Aghbari

Addresses: Department of Computer Science, University of Sharjah, P.O. Box 27272, Sharjah, UAE. ' Department of Computer Science, University of Sharjah, P.O. Box 27272, Sharjah, UAE

Abstract: We consider the problem of anomaly detection in data streams, which is the problem of extracting subsequences that do not match an expected behaviour. The main challenge for detecting anomalous subsequences from data streams in the existing techniques is to determine the lengths of the normal and anomalous subsequences. Therefore, creating a robust model for detecting the anomalous subsequences is of critical importance. In this paper, we propose an incremental algorithm based on the dynamic time warping technique to detect anomalous subsequences in data streams. The proposed algorithm is able to detect anomalous subsequences under relaxed length constrains of the normal and/or the anomalous subsequences. That is the proposed algorithm is able to detect variable length anomalous subsequences from among variable length normal sequences. The proposed robust model can be applied in areas such as system health monitoring, event detection in sensor networks, and detecting eco-system disturbances, etc. The cost of the proposed algorithm is linear with time and memory.

Keywords: anomaly detection; outliers; data streams; data mining; subsequence extraction; incremental algorithm; variable length subsequences; dynamic time warping; modelling.

DOI: 10.1504/IJIIDS.2012.047005

International Journal of Intelligent Information and Database Systems, 2012 Vol.6 No.3, pp.273 - 288

Received: 09 Feb 2011
Accepted: 17 Sep 2011
Published online: 16 Aug 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article