Collecting data from conflict zones by using social media and newspaper data

Summary

This innovative research project combines social media and newspaper data with more traditional information collected by governments and NGOs to predict forced migration movements in Iraq. The research team developed a methodology to extract information from text of open social media data to complement traditional data sources, since in conflict zones and regions with high poverty and political instability data collection via surveys and interviews often times is dangerous and challenging.  The study was conducted with data from the time period between January 2015 and December 2017.

Results

This innovative approach of incorporating open source data from newspaper and social media showed that the accuracy of the predictive estimates improved in comparison to research with only traditional data from governments and NGOs. The data proved as timely, uncontrolled – since everybody with a social media account can provide information -, and accessible, despite the danger surrounding the violent and armed conflicts in Iraq. Overall, this innovative data project demonstrates the additional insights that can be unlocked by complementing traditional data sources with new data such as from social media. Especially for contexts with limited and/or dangerous (survey) data access, this approach is a viable way to gain predictive estimations on forced migration movements and it can be replicated for other contexts as well.

(Image: © Muse Mohammed, IOM)