07 August 2014 by Dominique Grelet, Group Director of Strategic Projects and Global Offers, Bull
When crossing internal data segments with segments that are external to the company, new types of risks are created for which standard security management is not generally prepared. It is therefore necessary to manage security in a different way.
In addition to the classic principles for handling security (perimeter security, security in-depth) firstly, certain data segments may need encrypting and secondly security monitoring systems need to be deployed, in particular in relation to events that may reveal security breaches.
We are therefore going to use Big Data to strengthen security by identifying large amounts of important events, which considered one by one may appear innocuous, but taken as a whole will reveal a security breach (e.g. the opening and closing of a port over a short time scale, or a data stream to “exotic” servers).
To analyse this mass of events we use Big Data types of mechanisms, in particular algorithms that are applied to data to analyse weak signals. Big Data provides a holistic viewpoint of a whole set of events.