Qualitative and Artificial Intelligence-based Sentiment Analyses of Anti-LGBTI+ Hate Speech on Twitter in Turkey

Abstract

The aim of this study was to evaluate hate speech in Turkish LGBTI+-related tweets during a one-month period of artificial intelligence-based sentiment analyses. Turkish tweets related to LGBTI+, were retrieved using Python library Tweepy and were evaluated by sentiment analysis. The researchers then performed a qualitative analysis of the most frequently liked and retweeted tweets (n = 556). Sentiment analysis revealed that 69.5% of tweets were negative, 23.3% were neutral, and 7.2% were positive. The qualitative analysis was grouped under seven themes: LGBTI+ Club; Terrorism and Terrorist Organization Membership; Perversion, Illness, Immorality; Presence in History; Religious References; Insults; and Humiliation. The results of this study show that anti-LGBTI+ hate speech in Turkey is significant in terms of both quality and quantity. As LGBTI+ individuals are at risk for excess mental distress and disorders, it is important to understand the risks and other factors that ameliorate stress and contribute to mental health in social media. © 2023 Taylor & Francis Group, LLC.