Spatializing Syria's
YouTube War

Spatializing Syria's YouTube War Platform

authors: Violet Whitney, Michael Storm, Nadine Fattaleh
advisors: Laura Kurgan, Dare Brawley, Grga Basic
Center for Spatial Research

 

A platform that geo-locates YouTube videos from citizen journalists in the Syrian Civil War in Aleppo.

locative media, citizen journalism, conflict mapping, geo-fencing, location based media, context aware, location-aware computing

 
 

The rise of cellphones with video-cameras, as well as online video-hosting have changed the landscape of War documentation. Since January 2012, according to official YouTube figures, over a million videos of the Syrian Civil War have been uploaded, with hundreds of millions of views to date. Because of YouTube’s association with the rise of the war, it’s easy to see why it has been nicknamed “The YouTube War”.

 

Large News agencies have come to rely on YouTube as a primary source for ongoing events as dangers make it nearly impossible for western and outside Journalists to access these war zones. This puts increasing pressure on local and civilian journalists to document the events in Syria and distribute them to the world.

 
 video footage of the Syrian Civil War is shown through the window of a YouTube page

video footage of the Syrian Civil War is shown through the window of a YouTube page

 

However, with over a million citizen video uploads, there remains questions of credibility and bias as each YouTube channel hosts its own narrative and political agendas. In some instances, YouTube is a conduit for rebel groups to communicate with sponsors and spread videos of their accomplishments for funding.

While maps and satellite imagery give a broad yet reductive view of events in Syria, on-the-ground cellphone footage captures an intimate yet fragmented understanding. It becomes increasingly important to unpack the tension between these scales to understand the limitations of each. Especially as viewers around the world peer into the War from pages devoid of context. Where were these videos taken? Who filmed them?  

 
 map visualizations provide broad yet reductive view

map visualizations provide broad yet reductive view

 video provides an intimate understanding but is devoid of spatial context 

video provides an intimate understanding but is devoid of spatial context 

Our motivation was to create a platform and archive of mapped videos of the Syrian Civil War that will allow for a broad, zoomed out analysis of reporting on the war in Aleppo. By spatializing these videos, we hope this platform will become a tool for Journalists, researchers and other interested parties to sort and search through the YouTube dataset using a new method -one which is curated spatially.

By using Google Developers YouTube API we were able to extract archived and real-time metadata from the videos. Considering the large amount of content, a sample of video channels was collected from videos embedded in major western news organizations like The New York Times and Al Jazeera English, as well as videos used in analysis by Human Rights Watch. The intention was to use editorial expertise and human rights expertise as a measure authenticity and trustworthiness of various channels .

 videos are recorded as a specific location and could be sorted, visualized and understood spatially

videos are recorded as a specific location and could be sorted, visualized and understood spatially

This is the platform we developed which maps videos from the Halab News Network channel. We focused on this channel because the critical role it has played as a source vetted by Human Rights Watch. For those not familiar with Aleppo or the War, the Western half of Aleppo is an area held by the Assad government, while the eastern portion is held by Rebels, and a small portion of the north is occupied by Kurds. The gray boundaries represent the over 50 neighborhoods of Aleppo.

The circles represent the number of geotagged videos in each neighborhood in the Halab News Network channel. You might already notice that Halab News Network has recorded events mainly in rebel held territory. This can probably be attributed to the fact that HNN is affiliated with the resistance to Assad’s Government and the fact that the government has been bombing the rebel territory.

Screen Shot 2017-01-26 at 5.04.42 PM.png

By selecting any neighborhood, we can see all tagged videos of that neighborhood in chronological order starting from the most current. However, locating these videos was only possible because the strict naming convention used by civilian journalists who repeatedly use the neighborhood names in the video title descriptions. Given the varying levels of phonetic similarity between arabic characters especially as substitute characters are often used depending on available letters on the arabic user’s keyboard, it was important to understand alternative spellings and nicknames for neighborhoods.

As a native speaker, Nadine developed a list of alternative names for all the neighborhoods of Aleppo. We then developed a function that decides what neighborhood a video belongs to by matching its video title to possible neighborhood names.

 Citizen journalists in Syria use Neighborhood names in the YouTube video titles which allows for these videos to be placed spatially based on these neighborhood names.

Citizen journalists in Syria use Neighborhood names in the YouTube video titles which allows for these videos to be placed spatially based on these neighborhood names.

Parallel to the development of an interactive platform for the Halab News Network, we analyzed the metadata scraped from each channel in a grasshopper script to better understand the spatial and chronological histories of users in this space. On the screen is a chronological analysis of the HNN from 2013 onward. The left column has a line for every upload to the channel, the center column contains only geolocated videos. These numbers are represented as a bar graph on the right side of the visualization, divided by the date periods in landsat imagery discussed in the Change Mapping case study presented earlier.  

The advantage of this method of processing scraping channel metadata is that it can be quickly replicated across multiple channels. Here is the same information, reduced to all uploads on the left, and a bar graph of geolocated videos, compared across five channels.  

timelinegiffy.gif

Plotting the results of the geolocation data by neighborhood indicates the geographical territories of each group.  

The 16th Battalion is a rebel group operating in the northwest of the city; SHAHBA, HNN, and Aleppo Media center are all journalism organizations; and Syrian Civil Defence Aleppo, the local branch of the Syrian White Helmets. A quick overview of this analysis shows its relevance as both an analytical tool for investigating events in Aleppo as well as evidence of the processes of war, as users emerge and disappear over time.


While YouTube processes videos from a channel as a list, analysis of these channels shows spatial biases important to understanding the specific relevance of each channel to the larger conflict.

This project will continue with an archiving of several of the most complete Youtube Channels analysed so far; Aleppo Media Center, Halab News Network, and the Civil Defense Aleppo, know also as the White Helmets.  This, along with open source access to the methodology for processing and visualizing YouTube data will be hosted by the Center for Spatial Research. We believe that this new method of looking at YouTube videos will open new avenues of researching cities and conflicts, providing  new way of spatially understanding conflict in real time.