An artificial immune system approach to anomaly detection in multimedia ambient intelligence

Artificial Immune Systems are inspired by biological immune systems, and are characterized by interesting properties such as error tolerance, adaptation and self-monitoring. An area where they found wide application is anomaly detection in information systems, including intrusion detection. In this work we propose to extend the Artificial Immune System (AIS) paradigm from its typical application domain, computer system security, to ambient intelligence. AISs can be used to respond adaptively to real word anomalies in controlled environments. Here the counterpart of perceptual functions and detection capabilities can be provided by device intelligence, e.g. in terms of multimedia interpretation.