by Vedat Dogan (pictured), Insight at University College Cork

Did you know that the deaths caused by terrorist attacks have increased fivefold since the attacks on the United States on 11 September 2001? As terrorism activities are increasing each year worldwide, analysing the structure of a criminal network is becoming vital. In particular, identifying an individual’s role in a criminal network can help minimise casualties and understand the structure and potential duties of an individual inside the network. For instance, if we can identify the role of a person in a network at an early stage of a criminal network construction, the information can be used to warn counter-terrorism authorities, understand the possible influence of this person and prevent a possible future activity. Moreover, the information can be used to reshape the network after the death of an influential person on a network.

So, how can we identify the role of an individual in a criminal group? To identify the influence or role of a person in a criminal group/network, you need to consider a range of factors, such as the number of connections of the person, who the person is connected with, the background information, statistical network structure and many more. Even though AI presence comes with many more alarming concerns, one of the many things that it can do well is combine all this information, learn from it and make accurate predictions regarding this knowledge.

We are also using machine learning to solve problems in risk analysis and cyber-security.  To plan against attacks we must rely on estimated probabilities, our beliefs regarding the intentions of attackers, our beliefs about their beliefs regarding our intentions, and so on.  We are working on a framework for modelling and solving these problems using advanced AI algorithms.

I am working with Dr. Steven Prestwich and Prof. Barry O’Sullivan at the Insight Research Ireland Centre for Data Analytics, University College Cork, on a framework based on advanced AI algorithms for Diplomacy (AI4Diplomacy), including the prediction the role of an individual in criminal networks by extracting information from all the data mentioned above. It can be used by the authorities to understand the role of a person in a network to prevent criminal activities before they happen or to clarify an unknown person’s act in a group. Moreover, with AI algorithms, the amount of effect of characteristics or features to identify the role also can be explained.