- True identification; There is evidence of users creating fake profiles on purpose, to steal other’s identities or to engage in ‘cyber-bullying’. This limits the possibility of establishing reliable access to one’s online postings. Therefore, when collecting intelligence from open sources, the analyst must always be aware that even when a social media profile is created with a real name and surname, it does not necessarily belong to the person of interest. There is no ideal method to verify one’s online identity without confronting the person in question;
- Best practices in terms of online identification involve (a) cross-checking the data available from the social media profiles with the dataset created on the basis of the visa application or other existing information on file about that individual; and (b) remaining suspicious about the users’ perceived identity, especially when drawing conclusions;
- Language and cultural barriers; Cultural barriers and lack of knowledge, may prohibit security practitioners from understanding the specific context and true meaning of one’s postings. For example, without the proper knowledge about current trends in extremist ideology and propaganda it may be difficult to pinpoint suspicious Internet activity associated with terrorism;
- Best practices to address cultural challenges involve better training of security personnel and more diversity promoted among security practitioners;
- Governance of the data, lack of transparency in processing users’ personal information and insufficient information about the models that social media companies use for classifying users; Research has shown that Facebook underestimates the number of expats for certain nationalities, while considerably overestimating for others. Due to the lack of detailed documentation on how Facebook classifies expats, it is not possible to distinguish biases related to the selection and non-representativeness of the users or other noise and inconsistencies in the data;
- Samples of social media data are not typically representative of an entire population, since not everyone uses social media and levels of penetration vary in different locations. Frequent users may choose not to provide information on past and current location; It is also very difficult to verify whether changes in location are accurate and the use base of social media providers constantly undergoes changes. Even with these limitations, when events occur or a situation is deteriorating in a region, online conversations do emerge and these discussions, even if they are not specifically about migration, capture indirect indicators of one or more migration variables;
- Perceptions can be measured in different ways. Three that are important in the context of migration are tone (sentiment), stance (position – for or against), and emotion;
- Location cannot always be easily established. While newspaper articles generally contain the locations that the articles refer to, social media posts are less consistent;
- Best practices to address the location identification challenge are by: analysing geolocated posts, collecting the location of the user posting, and/or collecting mentions of locations in the post text;
- In some cases (e.g. there is evidence of this applying to content in Arabic) sentiment and tweet volume are indirect indicators of movement associated with conflict factors.
Cesare, N., Lee, H., McCormick, T., Spiro, E., & Zagheni, E. (2018). Promises and pitfalls of using digital traces for demographic research. Demography, 55(5), 1979–1999.
Karasek, P. (2010). Social Media Intelligence as a Tool for Immigration and National Security Purposes. Internal Security Review Vol. 19 No. 8, pp. 405-415.
Singh, L., Wahedi, L., Wang, Y., Wei, Y., Kirov, C. et. al (2019). Blending Noisy Social Media Signals with Traditional Movement Variables to Predict Forced Migration. KDD ’19: Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, pp. 1975–1983.
Spyratos, S., Vespe, M., Natale, F., Ingmar, W., Zagheni, E., & Rango, M. (2018). Migration Data using Social Media: a European Perspective. Publications Office of the European Union, EUR 29273 EN, ISBN 978-92-79-87989-0.