Applying Cognitive Technology for Anti-Terrorism Predictive Threat Detection


By Gary Olson, Co-Founder, CTO, MediaMaestro  |  February 1, 2018


In today’s world, terrorism and cause-inspired violence against innocent people fill the news media and has put the public, law enforcement and the governments of the world on edge.  Media Maestro is using cognitive computing coupled with advanced media workflow orchestration to create a Cognitive Intelligence solution for predictive analysis to help prevent these acts of terrorism, attacks and violent situations or threats.  

Public and private security organizations have the capability to capture vast amounts of video, imagery and big data from sensors, radar, social media and many other sources.  Each of these sources are collected in isolated data pools using proprietary tools to mine and analyze each data type.  Sometimes these tools are specific to only one data format.  In many cases the analysts are in geographically separated locations while reporting to the same manager and/or decision maker.  

MediaMaestro considers each of these data pools as a source, and through analysis aggregates them creating a virtual data lake of all the data sources, formats and types. MediaMaestro analyzes all of this in a single application and intuitive user interface.  The workflow orchestration engine manages the cognitive processes as the data is curated.  The cognitive technology has the criteria and thresholds that discover matching data points or behavioral patterns indicating a potential threat condition and triggers an event alerting an analyst to review actionable information.

There are a number of consumer facing applications where cognitive computing is in use. This includes various recognition technologies uses such as facial, character, object and activity detection.  In many of these instances, a dedicated data pool is created.  In some cases, where multiple cognitive processes are used, a data lake is created.  Some of these data pools may be unstructured making them less useful.

MediaMaestro is the next layer in the stack.  Cognitive Intelligence is the next generation of the Intelligence Cycle (Direction, Collection, Processing, Analysis, Dissemination, and Evaluation & Feedback).  Currently, each phase of the intelligence cycle is handled separately by different departments, personnel and tools.  This new technology augments each phase within a single application.  This creates a more efficient and effective method of analyzing this considerable amount of data into actionable information.

Intelligence cycleMediaMaestro’s workflow orchestration engine, described in above image, manages all of the cognitive processes. As the system operates, the orchestration engine engages and disengages cognitive processes dynamically as these apply to each of the different data sources and types.  This cognitive set of processes generates new knowledge, which is able to identify threats and situations that match the threat criteria data points.  It evaluates the assets on a changing threshold scale to determine if there is actionable intelligence.  This is MediaMaestro’s predictive process that enables a decision maker to initiate preventative action. 

The amount of security video being generated by satellite, surveillance, body and car cameras, postings on social media and UAS (Unmanned Aerial Systems) plus the data from sensors, radar, unstructured and structured sources continues to grow and is drowning analysts. This is creating an overwhelming amount of video and big data that needs analysis and has produced a “Firehose in a Dixie Cup” effect of almost too much video and data.  National Security, Public Safety and Private Security organizations that have the responsibility to prevent attacks or threat incidents have realized the inability of human analysts alone to be able to efficiently and effectively identify these potential threats or situational awareness concerns in a predictive and preventative way. MediaMaestro enhances the analyst’s ability to produce actionable information.

There are numerous technologies collecting these data sets and creating individual data pools.  These data pools are segregated from each other with different tool sets applied to index and search.  MediaMaestro uses these segregated data pools and creates a virtual data lake curating all the different data types and formats through a unified set of cognitive analytic processes to assess the value of the data against a set of threat criteria and deterministic dynamic threshold analysis to evaluate the level of threat and then trigger a recommendation for action.

This is how MediaMaestro has created cognitive intelligence as an outgrowth of cognitive computing.  Cognitive Intelligence is the application of an analytic set of processes integrated with the cognitive engine and aggregating the output of multiple cognitive processes into a new analytic layer refining the output with greater accuracy and efficiency.

Applying the next generation of new knowledge from an existing knowledge base is the core of cognitive computing.  Using pattern analysis and relational data comparison to evaluate the efficacy of the new knowledge and applying it to a set of deterministic conditions is cognitive intelligence.

Assessing seemingly disassociated data points across multiple diverse data sets is the core of cognitive intelligence.  If different applications show similar intent based on information captured from a video camera, data sensor, or a social media posting, other data tools do not detect this similarity or correlate it to a broader or potential threat condition.   Using our new techniques in cognitive analysis, MediaMaestro assesses each of the data points and using inference based intelligence, deterministically produces resultant data that matches pre-determined threat criteria.  Once the threat is validated, preventative action can be taken to avoid an attack or incident. 


Media Maestro

MediaMaestro evaluates and identifies across users within the system commonalities and alerts a senior decision maker that more than one analyst is potentially looking at the same or similar asset, activity or person that is considered a threat.  The system’s Connect the Dots feature enables the senior decision maker to bring the analysts together or assign the event to one of them.    

MediaMaestro is the next generation of Predictive Analysis of Situational Conditions and Threat Detection to enable preventative actions that will mitigate attacks or incidents threatening National Security and Public Safety.

For More Information, Contact Gary Olson gary@gtopgroup.com