Bahavior-based Security

Contacts: Arik Messerman, Tarik Mustafic, Joël Chinnow, Seyit Ahmet Camtepe


The behavior of system parts or people in an environment in the form of state descriptions as well as transitions contains a wealth of information and data that, if they are evaluated efficiently and applied by analogy, are helpful for the investigation as well as prevention of unwanted occurrences. One specific focus lies on the fact that the behavior can vary over time. Another challenge is the management of large amounts of data that represent the behavior. This implies among others appropriate measures for data reduction and abstraction of behavioral characteristics.

A special case of behavior analysis is located in the field of so-called behavioral biometrics. This research direction aims the analysis of behavioral characteristics of individuals for identification and verification purposes. Especially in the analysis of personal behavior patterns, in contrast to the physiological biometrics where the most immutable patterns of a person (e.g., a fingerprint) are used, high behavioral variations have to be compensated. Those variations are subject to natural external influences such as the stress, joy, or haste of a person, etc.

The CC Security increasingly focuses the analysis of behavior of certain persons during their interaction with computerized systems, thus developing solutions that can be used for transparent and continuous verification of persons based on their behavior during the interaction in real time.

A further goal of CC Security is the analysis of the behavior of a user while interacting with a mobile device. Today's smartphones have sufficient resources and computing power in order to be used to identify a person continuously by using behavioural biometrics. In this way, a mobile phone can be considered as a biometric sensor, which is almost always carried with them.