 |
| Credit: http://scottwatson.co |
It has been argued that in order for SOCMINT to be effective, two tests have to be passed – a solid methodology of collection, evidence, verification, understanding and application; and ability to manage a moral hazard which can arise from using this type of collection (Omand, 2012).
Several benefits of collecting the social media intelligence have been pointed out: collection of such data can help with identifying pathways into radicalisation, or, serve as indicators of violence. Moreover, monitoring long term social and political causes of terrorism can help with formation of counter-terrorism strategy. In addition, social media networking can help with tension monitoring.
Few years ago, a ‘tension engine’ component of Cardiff Online Social Media Observatory (COSMOS) has been developed. It monitors social media data streams for signs of high tension, which can be analysed to identify deviation from the norm (Williams et al, 2013). As a result, a ‘Collaborative Algorithm Design' has been introduced, which draws on different sociological concepts. Spikes in tension are important in predicting social disorder, but it has been pointed out that sentiment alone is not enough for tension detection (Williams et al, 2013).
 |
| Credit: http://www.shineinfotech.com |
On the tactical level, social media intelligence can identify criminal activity, early warnings of social disorder and help with understanding and reacting to public concerns. By using new developed geo-location techniques, the location identification of criminal activity can speed up the emergency response (Omand, 2012). Social media intelligence can be used to determine behavioural patterns applying to certain groups or individuals and predict future trends (Ivan et al, 2013).
However,
there are also challenges of SOCMINT collection. There has been very little
social research done in order to develop a credible approach to robustly sample
social media data sets. More research is needed on social media data sets that
would acknowledge the sampling frame applied and how this might limit or bias
the drawn results (Omand, 2012). It has to be stressed that in order to
analyse data correctly, inclusion of human factor is very important, as the
data can be easily misinterpreted. In addition, large amount of data can create
‘noise’, and overloading and oversampling errors can affect the intelligence
cycle and cause misinterpretations - leading to wrong decisions (Ivan et al,
2013). Another risk Omand et al. defined is an ‘observational effect’. Research has found that people behave
differently when they are aware of being observed, and this could also have a
negative effect on the intelligence analysis itself.
Observational effect – falsification of information
Data collected online can be discredited by observational effect and users might not provide accurate information about themselves, research (2014) has found. Some types of information are more likely to be falsified than other. A better understanding of the tendency of Internet users of when and why they change their online behaviour in response to online surveillance can help pinpoint especially problematic areas for the validity of open source (OSINT) or social media intelligence methods.
A survey conducted in 2014
revealed,
the more negative peoples’ attitude is
towards governmental online
surveillance,
the more likely they are to provide
false information (Saskia Bayerl,
Akhgar, 2015).
Hence, properly legitimised online surveillance could reduce
distrust in law-enforcement agencies and pressures towards information
falsifications. Law enforcement agencies have to consider falsification of
information when analysing the data for intelligence purposes, and account for consequences of their
decisions based on these collection methods (Saskia Bayerl, Akhgar, 2015).
Corroboration of information, such as identification of links between profiles
of a single user and further collection might be necessary to protect the
integrity of force intelligence units.
References:
Omand, D., Bartlett, J.,Miller, C. (2012). Introducing Social Media Intelligence (SOCMINT). Intelligence and National Security, 27 (6), 801-823.
Omand, D., Bartlett, J.,Miller, C. (2012). Introducing Social Media Intelligence (SOCMINT). Intelligence and National Security, 27 (6), 801-823.
van de Velde, B., Meijer, A., Homburg, V. (2014). Police message diffusion on Twitter: analysing the reach of social media communications. Behaviour and Information Technology, 34 (1), 4-16.
Saskia Bayerl, P., & AKHGAR, B. (2015). Surveillance
and Falsification Implications for Open Source Intelligence Investigations. Communications of the ACM, 58(8), 62-69.
doi:10.1145/2699410.
Williams et al. (2012). Policing cyber-neighbourhoods:
tension monitoring and social media networks. Policing and Society: An International Journal of Research and Policy,
23 (4), 461-481.