Worldstack | A.I is disrupting ISR and Business Intelligence
7975
post-template-default,single,single-post,postid-7975,single-format-standard,ajax_fade,page_not_loaded,,qode-theme-ver-2.9,wpb-js-composer js-comp-ver-4.12,vc_responsive

Blog

A.I is disrupting ISR and Business Intelligence

  |   Open Source Intelligence, OSINT, Uncategorized

What is ISR?

 

Intelligence, Surveillance and Reconnaissance (ISR) is a concept that has its roots in the military. It refers to the capability of acquiring and processing information and disseminating intelligence. The business equivalent is Business Intelligence.

 

 

Challenges in ISR

 

1. Lots of data, at speed

ISR data can come from a variety of sources including satellites, manned aircraft, unmanned aircraft (such as drones), ground or sea based systems, human intelligence teams and online (cyber). In many cases, this data is unstructured.

 

Some of these sources, such as unmanned aircraft and online generate a lot of data. Unmanned aircraft platforms are capable of high definition full motion video. In the online space, Facebook alone generates 250 million posts an hour.

 

2. Availability of analysts

Someone has to analyse that data to produce intelligence. There is far more data than there are analysts to process it.

 

3. Time to target

The time it takes to go from data to decisions is critical. It can mean the difference between an opportunity to act to prevent or take advantage of an event, or not.

 

 

A.I Disruption

 

1. Enhanced search

Use of natural language processing to improve how search is conducted over large datasets. With a base set of words as a search target, the system is able to find words that are similar or mean the same thing as the base set, reducing the time it takes to formulate a search and improving the search results.

 

2. Identity and entity resolution

Use of natural language processing and machine learning to perform entity resolution. For example, collecting information on a person of interest based on known attributes and then populating that entity with other discovered attributes, to form a complete picture of that person from available data. Another important feature of entity resolution is de-anonymisation –  the ability to link an online persona to a real-world identity.

 

3. Text classification

Application of machine learning to large streaming data sets such as surface, deep and dark webs to identify events of interest such as changes in sentiment or changes in human behaviour or to rank search results according to user interest.

 

4. Image, voice and video classification

Use of computer vision techniques over large image, voice and video collections to perform tasks like detecting threats, identifying known persons of interest or changes to environment.

 

 

Bringing it all together

The combination of a data set that is rapidly increasing in volume, velocity and variety and the limitations on human resources to process that data and generate actionable intelligence has created a significant opportunity for the application of artificial intelligence.

 

By applying computer vision, machine learning and natural language processing techniques to the increasing volume of unstructured data, organisations can make the most effective and efficient use of limited analyst resources to generate actionable intelligence that is relevant and timely.

 

 

About WorldStack

WorldStack is a private security company that specialises in open source intelligence. By applying leading edge artificial intelligence techniques to unstructured data, we are unlocking the secrets of the surface, deep and dark web, at scale.

 

 

About the author

Dan Holman is a founder and CEO of WorldStack. He is an experienced IT project executive and cyber security consultant with government, private sector and international experience.