Categorization of Maritime Anomalies for Notification and Alerting Purpose

PDF

Authors
  1. Roy, J.
  2. Davenport, M.
Corporate Authors
Defence R&D Canada - Valcartier, Valcartier QUE (CAN)
Abstract
Automated anomaly detection systems in the maritime domain typically generate large numbers of unorganized alerts of various types which must be reviewed by human analysts. When the volume of reports becomes very high, and when the reports cannot be structured and/or prioritized, there is a greater risk that analysts/subscribers will reject the whole data stream. Efficient mechanisms are thus needed to make the reports more palatable to the analysts and/or the partner agencies on a network who may subscribe to an anomaly-reporting service. Such mechanisms require a categorization of the anomalies that is well-matched to the implicit operational models of the end-users. This paper presents such a categorization, based on a taxonomy of the maritime situational facts involved in anomaly detection identified and validated through knowledge acquisition sessions with experts. The paper begins with a. review of domain challenges, knowledge-based (expert) systems, and the knowledge engineering process that was used. Highlights of results from knowledge acquisition sessions with maritime domain experts are presented; these results are the foundation of a proposed taxonomy of maritime situational facts of interest. This taxonomy is then used as a mechanism to categorize maritime anomalies.
Report Number
DRDC-VALCARTIER-SL-2009-394 — Scientific Literature
Date of publication
15 Sep 2009
Number of Pages
10
DSTKIM No
CA033343
CANDIS No
532651
Format(s):
Document Image stored on Optical Disk

Permanent link

Document 1 of 1

Date modified: