V11-33

Systems employed during surveillance activities might be enhanced using anomaly detection capabilities. Such capabilities can be employed to highlight those situations, events or objects that need operator's attention, reducing, thus, their cognitive load and reaction time. Early detection of such situations provides critical time to take appropriate action with, possibly before potential problems occur. However, the detection of such conflict situations or general anomalous behavior in surveillance data is a complex analytical task that normally cannot be solved using purely visual analysis or purely automatic computational methods. On the one hand, the success of purely visual analysis methods for area surveillance often depend on factors such as the amount of sensor data that needs to be monitored, time constraints, or even operators' cognitive load and level of fatigue. On the other hand, current automatic anomaly detection solutions normally present high false alarm rates when dealing with complex situations. The high number of false alarms can become a nuisance for operators, who might react by turning anomaly detection capabilities off. Some researchers dispute the use of fully autonomous discovery systems in real-world settings, highlighting the need of including human knowledge in the discovery process. Most of the published work on anomaly detection focuses on the technological aspects: new and combinations of methods, additional improvements of existing methods, reduction of false alarms, correlations among alarms, etc. Publications regarding the use of anomaly detection methods in real environments, or human factors studies regarding anomaly detection, are scarce. In order to find optimal combinations of human expert knowledge and computational methods for anomaly detection, it is important to investigate how the surveillance of sea areas is carried out. This domain is suitable for the study of finding optimal combinations of expert knowledge and computational methods, since it fulfils the characteristics of many data-intensive domains – large amounts of multivariate data, the need for operator support to solve complex problems, the need for situation awareness to promote effective decision-making etc. Knowledge of how the analysis of traffic data is carried out in a daily-basis can be used to propose how to support such processes using data mining and visualization methods.

According to the passage, which of the following does the author consider a criterion for selecting a field of research in automation optimisation?

A. Enhancement of safety aspects
B. Inadequate research in the field
C. Number of false alarms generated
D. Complexity of analysis
E. Requirement of human knowledge