Tweeting terrorism - following extremists in cyber space

THE SPREAD OF RADICAL AGENDAS

An interesting sub-field of counter terrorism in the post 9/11 world has been the study of radicalization. It is believed that terrorists are influenced through a process of incitement, which begins with

According to a report from the New York City police Department, the Internet is a driver and enabler for the process of radicalization in the west. This is quite intuitive; the internet provides a certain amount of anonymity and more importantly – a virtual meeting place for those with radical ideas. These meeting places carry even more weight in the west, where radical ideologies aren’t taught at schools or places of worship, but rather have to be sought out by the individual.

Since online social networks allow the spread of idea to a large variety of people in a short amount of time, the danger of incitement becomes even stronger. I would be interested in researching the spread of violent and radical ideas on English speaking Twitter accounts. English language twitter accounts are geared towards potential terrorists in the west. What we refer to as "home grown" terrorists live inside the own culture they are trying to attack, they decide to commit their acts not due to the culture they live in or their family upbringing, but rather due to radicalization through like-minded peers and social networks.

THE TWITTER NETWORK

In recent years terror network communications have become less hierarchal and more flat. There has been a shift from few password protected forums to public online communications often maintained by supporters.

I would choose to focus on the most open of tools, and one that has been used frequently by terror organizations – Twitter. The huge advantage of Twitter for researches is that most information is public. Unless a user chooses to privatize his or her account (rare), anyone can see who one follows and who is followed by them, as well as the content of their posts.

KEY CONCEPTS

One key concept in networks that can help us analyze the spread of extreme ideas is Homophily, the tendency of people to associate with others like themselves.[1] This concept is of strategic importance since it leads us to assume that people in the many of the connections are between likeminded people with similar values or beliefs.

Those following terrorist organizations on Twitter are more likely to be influenced by the ideas and distribute them. In the study of diseases, those who contribute to the spread of the epidemic are the susceptible population. Similarly, with the spread of ideas we can assume that those who “carry” the extreme ideology are predisposed to it. What is especially interesting in this case is that as opposed to the study of health, where we know the conditions that help carry a disease ( poor nutrition or hygenie ), the vulnerable human factors that carry ideology are not always known. Therefore, instead of searching for networks in impoverished areas and follow the disease spread, we will instead follow the disease and learn from it about the attributes.

METHODOLOGY

As mentioned, the huge advantage of twitter is that all information is readily available. I would therefore focus on a few accounts associated with terror organizations and assess their followers.

ATTRIBUTES

I would consider the following attribute for each follower:

1. Number of followers:

a. Below 50 – low influence

b. Between 50-250 – medium influence

c. Over 250 – high influence

The number of followers represent how many people have actively chosen to subscribe to this user’s fee, therefore hey are a good measure of influence.

2. Number following:

a. Over 50 – mildly active

b. Between 50-250 – moderately active

c. Over 250 – extremely active

It is important to note that not all the accounts a chosen user follows will necessarily relate to extremism. However, it is safe to assume that the more followers one has, the more active they are on the network in general and are therefore a good measure for the spread of ideas.

Location:

Many people choose to list location on their Twitter account. Even if location is not listed, it is not impossible to use IP addresses to attempt to track down a general area from which a user is tweeting. The location attribute is extremely important because it allows us to observe whether clusters are formed in certain areas. If we see that many followers of an account are centralized in one are, law enforcement can try to contact a religious leader in that region and warn them about the radicalized population. Religious leaders often work with law enforcement to attempt to de-radicalize members of their community that have been identified as potentially harmful.

TIES

I assume this network will be decentralized and scale-free as is often the case with internet connections.

Inedegree and outdegree ties of nodes will be of extreme importance.

1. a. Indegree centrality – how many people follow a certain node?

b. b. Outdegree centrality– how many people does a certain node follow?

Though the strength of these ties cannot be measured, this measurement offers three great advantages for the researcher:

1. Simplicity – the ties are very clear: you wither follow someone on twitter or you don’t. These are not subject to possible manipulations in surveys

2. Identifying Closeness – this is a huge advantage in the field of counter terrorism since it allows us to identify who is connected to people and can reach the most users in the least amount of steps.

3. Density – identifying whether a network is dense or not can have great implications in the field of Counter Terrorism. It will allow us to assume how well connected and organized a movement is, or whether it is made of loose nodes that are not necessarily all communicating with each other frequently.

ADVANCED RESEARCH

Advanced research can also measure tie strength by taking into account which users are communicating with each other. Once again, the advantage of this method is that all communications on Twitter are public, though the collection of data would be time consuming.

Collecting data on a number of user interactions would allow us to re-categorize and examine tie strengths: Which users communicated more than once? Which ones communicate on a weekly or monthly basis? In addition, directed and undirected closeness can show us who receives more messages (can imply that this person is a central source of information) and who sends more messages (this person is more eager to collect information)

CAVEATS

It is important to remember that this research, like many, has some caveats that must be taken into account. Most importantly, it is important to recognize that not all followers of terrorist Twitter accounts are necessarily extremists. In addition, even if one individual appears to be well connected, it does not mean this is grounds for action for law enforcement. However, as I mentioned, we have much to learn regarding density, location clusters and closeness from information that is 100% public, it just needs to be sorted!

---

[1] Valenta, Thomas, Social Networks and Health: Models, Methods, and Applications, Oxford Press 2010