Greek-Turkish border in March 2020 – Media Mining System for Migration Monitoring

Greek-Turkish border in March 2020 – Media Mining System for Migration Monitoring


This report focuses on the demonstration of some of the capabilities of SAIL LABS Media Mining System (MMS) through investigating the communication discourse regarding the migration crisis at the Greek-Turkish border. The aim is to show the value of having an ability to gather and process large volume of open source data in the context of a government institution to maintain situational awareness.

To access the full report, please contact us through the request form below.

This is a use case of the SAIL LABS OSINT tool suite. This work is compiled only on freely available information in context of the current issues pertaining to the EU/Turkish border contact line. This data includes TV, Radio, Web and Social Media information that was collected between the 1st of March 2020 and the 16th of March 2020. This data was collected and processed by the SAIL LABS Media Mining System and covers 32 languages among which the primary EU languages as well as Turkish, Arabic, Pashto and Urdu, which have high relevance in context of this particular migration issue. The intent is to showcase the value of having such information gathering and processing ability in the context of a government institution to maintain situational awareness and situational understanding. In this paper we only look at some details, so this is by no means a complete picture.

For brief reference by the BBC:
Following a deadly attack on Turkish troops by Syrian government forces in northern Syria, Turkey opened its borders with Greece to let the migrants pass the respective border at the end of February 2020. Turkey is hosting 3.7 million Syrian refugees, as well as migrants from other countries such as Afghanistan – but had previously stopped them from leaving for Europe under an aid-linked deal with the EU.

Problem: In a situation like the one at Greek-Turkish border, it is crucial to examine the communication discourse. Therefore, it is vital to investigate what do refugees gathering at Greek-Turkish border talk about, which media do they consume and share and if they are encouraged by particular actors.

Objective: Do the content analysis of open source refugees’ communication, identify relevant actors and assess the situation from the social media perspective.

Solution: The Heat Map provides information about the locations associated with social media posts. It features the Direct Coordinates (coordinates available from posts directly if GPS information is available) and/or the Detected Locations (locations detected in the text, the so-called Named-Entities or NEDs).

The bigger the area and the more intense the red color is, the highest density of the social media posts.

When we look at the area of the Greek-Turkish border, it is possible to spot clusters of social media posts being created in Edirne, Uzunköprü and Ipsala. When the area of the border is closely examined, it is possible to investigate the tweets regarding migration. If the geofencing is used, the tweets in the area will be collected in advance. Media Mining Client allows to investigate the meta-data of the twitter posts, who posted and retweeted the tweets and which users they are connected with.

Picture 1: The Heat Map displaying the areas of social media activity close to the Greek-Turkish border
Picture 2: Close detail of the tweets close at the border

When the social media content of the area is investigated, the system informs us about the location of the tweets (if the GPS information is available), the top tweets created in the area, the potential reach of the posts created, the hashtags social media users used and the network graph of their relations to each other.

It is possible to observe which profiles post images and messages in the area in real time. Afterwards, these social media profiles can be investigated in detail. This allows to find out what do the people in the area talk about, what media they consume and share or if they are being encouraged by misinformation.

Picture 4: Tweet details

As can be seen in Picture 4, the social media profile posts from the area and shared a picture with people walking besides the road and gathering at the border. The post could have been seen by more than 17 thousand social media users.

The Relations Graph in Picture 5 shows which users either post from the area of Turkish-Greece border or mention this location in their communication. We can identify the clusters regarding this topic and further investigate them. The Key Opinion Leaders (Picture 6) chart informs us about the most influential social media profiles sharing content with regards to this area and potentially influence the social media discourse.

Naturally, also the attitudes towards migrants differ in the population and social media users.

It is of high importance for any government to find out what do the citizens think about such a situation. There is an example of such a tweet below.

Picture 7: Example of a tweet Source: Twitter


To access the full report, please contact us through the request form below.

To access the full report and read all the chapters, please contact us through the request form below.