Conflictos armados y la construcción de narrativas a través de Twitter. El caso de la guerra entre Armenia y Azerbaiyán

Abstract

This article analyzes Twitter messages, in Spanish and English, during the development of the last conflict between Armenia and Azerbaijan over the historical Nagorno-Karabakh region. This conflict, along with the issues about Abkhazia and South Ossetia, is one of the most important crises in the South Caucasus. Two algorithms designed for processing large volumes of information have been used, namely LDA (unsupervised) and SVM (supervised). Based on the framing theory, a conclusion has been reached that both audiences are mostly positioned with the Armenian stance. The article also shows that the messages focus on other issues different from the war, such as Turkey's role, the Israeli government's responsibility for saling weapons to Azerbaijan's army, or the religious orientation of both countries. The results show that humanitarian explanations are rarely used by both audiences whose messages are rather focused on conflicting elements. In short, this work does not only seek to identify the elements of the public debate around conflict but it also highlights the potential of computer science techniques in political communication studies. (English) [ABSTRACT FROM AUTHOR]