Syrian Archive created a list of incidents based off publicly available research by the United Nations Independent International Commission of Inquiry on the Syrian Arab Republic, reports from many Syrian and International civil society organisations as well as journalists, activists, lawyers and individuals claiming that the attacks were conducted by the Russian air force. We have used flight observation data of aircrafts flying from Hmeymim airbase into the impact sites to corroborate with the mentioned claims. We were not able to do this corroboration for all incidents in this database. We are releasing this database online to allow for collaborative verification and corroboration for each incident that hasn’t been fully investigated.

Syrian Archive staff sifted through its existing database of over 2.5 million videos to identify documentation of targeting civilians and civilian infrastructure as a result of alleged Russian airstrikes. Relevant videos were located and verified using the Syrian Archive Digital Evidence Workflow. Metadata was extracted and standardised using an automated process and metadata scheme. Through undertaking this process, incidents not included on any of the human rights reports that were published by different organisations were uncovered.

For each of the 1418 incidents identified, a detailed process of discovery was initiated. This process necessitated:

Extracting units from the Syrian Archive database about specific incidents using keywords, dates and locations; Extracted units through specific geo-based areas; Using content discovery tools for social media content such as; Using computer vision tools to extract videos showing the use of cluster munitions.

Through adding context to data, Syrian Archive transformed digital content units into useful and searchable information with querying capabilities.

All content used in the building of this collection is open source and publicly derived, processed and published. The database places incidents on a map and allows data to be searched by incidents, keyword, and related video content.

While comprehensive in scope, there exist gaps in the collection. Some incidents are heavily documented while others are not. Continual updates of observations and incidents are planned, as well as inclusion of data from additional sources and publishing platforms.