Social networking and intelligence gathering
Despite our increasing reliance on technology to advance security capabilities, we still must rely on human intelligence to assess the significance of data. James Turner has the story.
Facebook is one in a class of websites referred to as a social networking site. A social networking website is one where subscribers are able to customise a web portal and create a schema which reflects their interests, personal information, and a variety of means to communicate via the site with their friends and associates.
It’s all about collaboration and multi-directional communication: not just the web site pumping out information like a tickertape, but the readers being empowered to interact with the site and with each other. This is the concept behind Web 2.0, and it is a compelling concept for humans; who are by definition, social animals. We place great meaning on relationships.
Most of the social networking sites serve different collaborative purposes, but the usual outcome is that you, as a subscriber, have created a page about yourself and a list of your known associates.
Data mining
The salient point is quite obvious – no matter how you try to construct your privacy settings, your details and interactions are in the websites’ database. This information is ripe for data mining.
So, if I was running an intelligence agency, I’d support a range of social networking sites and let people show me who they were connected to. It would save me years of legwork in intelligence gathering.
The problem is that data mining does not show the significance of the relationship between items of data. There may be a link, but what does it mean? A report for the US Congress: "Data Mining and Homeland Security: An Overview" (January 18, 2007) states that a limitation of data mining as a terrorist detection tool is the nature of casual relationships between data.
The examples given in the report relate to air travel tickets booked at short notice, and how these could be due to a hobby, an ill family member, the nature of a job, and not necessarily indicative of an impending terrorist attack.
Human intelligence
We can only do so much with technology. At a certain point we cannot escape the need to have human intelligence assess the nature and significance of the relationships between identities.
If we look at the example of the social networking phenomenon, the idea of variable significance makes sense. Two people may be linked, but they may barely know each other. Conversely, they may know each other very well, but their relationship on the social networking site may look distant or even non-existent. It takes human analysis to determine the significance of the matter.
Several years ago I learned an important distinction from a CIO: he told me that technology never solved a problem. Overcoming my surprise, I asked him what he meant and he pointed out that technology merely automates a process, and the process is what addresses the problem. People are required to create the process, refine the process, and support the technology which automates the process.
Of course, we need a balance of technology and human resources for intelligence gathering and intelligence assessments. The computers are brilliant at number crunching, but there are factors of significance which they may never be able to understand.
We can now build a computer which can outplay humans at chess. But how many hundreds of millions of dollars did it take to develop a machine which could calculate the millions of nuisances in the changing values of the pieces on an eight by eight board? Humans are still light years ahead in analysing the subtleties of life.
Filter their view
But humans have preconceptions which filter their view of the events with irrelevant meanings. Consequently, different humans will put different meanings on the same observation (classically termed as observer bias). So how do we program the humans and, at the very least, retain their superior ability to see significance in the most subtle of relationships?
Without addressing this question, we condemn ourselves to living in a restricted construct of the world.
About the author: James Turner is an advisor with IBRS, an Australian company that provides research and advice to IT and business managers in Australasian organisations. James specialises in the IT security sector. www.ibrs.com.au