Analyze Hate Contents on Sinhala Tweets using an Ensemble Method

As social media grew in popularity among the general public, content and opinion sharing has become rapid and convenient. The majority of users rely on social media for news and trust the content shared by the network. Some individuals, both purposefully and unintentionally, disseminate hate content and instill hatred in their readers. Even in Sri Lanka, the spread of hate propaganda on social media has resulted in communal discord and a variety of concerns. Only a limited number of research studies have been conducted to analyze the hate content written in Sinhala. This work investigates a mechanism for detecting hate content typed in Sinhala language and posted on Twitter. The proposed supervised mechanism is an ensemble method that selects the most accurate result from different models. 63% of accuracy, 58% of F1 Score, 61% of Precision and 58% of Recall were achieved when predicting hate content.