Shifting Policies in Conflict Arenas: A Cosine Similarity and Text Mining Analysis of Turkey’s Syria Policy, 2012-2016

Abstract

Turkish policy towards the Syrian civil war, as operationalized in relation to the implementation of no-fly zones, safe zones or buffer zones, has been the subject of much debate among scholars. As the number of foreign states acting in Syria has steadily increased since the onset of the crisis, Turkish policies have similarly shifted. In order to make sense of Turkey’s actions and reactions in the first five years of the Syrian civil war, this article attempts to draw lessons from quantitative methods and methodologies such as text mining, cosine similarity and cosine normalization of content from the Anadolu Agency (AA), a Turkish state-owned press. These methodologies are utilized in support of content analysis and qualitative analysis that hindsight allows. In doing so, we are able to show that these seemingly inexplicable shifts may adhere to a logic and, in some cases, could have been anticipated. Utilizing such methodologies therefore offers a potentially significant contribution to the literature by defining politically feasible outcomes related to foreign or domestic policies.