AI and Machine Learning
Defending against Complex Hybrid Threats
Social Network Analysis
Situational Awareness can depend now on space-based sensors, capable of providing data and imagery at high resolution which when combined with other sources of information, including all of the information available on the Internet, and population data organized by location, ad infinitum, can result in a big graph that resides in memory. Such graphs can be large. AI/ML tools like Reservoir Labs’ ENSIGN can identify potential anomalies and pass them for inspection to a human analyst. The role of AI in such a scenario is recognizing anything odd, behavior that does not fit a known pattern.
Humans can assist and control this process. Human intuition cannot compete with automation but given access to the right platform, a data scientist can probe data at a terabyte-scale. Meaningful behaviors can be described by Cypher queries and once discovered by digital forensics can then be referred to a Situational Awareness team, for example, to be incorporated into automated ML-type queries against a continually updated graph held in the memory of an operational system, perhaps a scale-out system, capable of petabyte dataset support.
Interconnected social networks are natural candidates for graph analytics. A significant portion of Internet traffic on social media platforms is misinformation deliberately spread and amplified by bad actors on the web and bots. A forensic exploration of social network traffic can highlight key influencers and opinion shapers. By following the supply chain for misinformation back to its source, a data analyst can possibly identify the instigator, or adversary, responsible.