abstract
- With the exponential growth of social media platforms, there has been a facilitated global message dissemination and diverse information exchange in real time. However, this communicative diversity can also nurture hate speech due to a myriad of opinions. This research focuses on detecting hate speech during the 2022 national strike in Ecuador, where derogatory remarks were directed against indigenous protesters on Twitter. By implementing a system in Python that employs the K-Nearest Neighbors (KNN) Algorithm in tandem with a dual-check method based on identifying words commonly associated with hate towards indigenous people, tweets are categorized into ¿hate speech¿ and ¿non-hate speech¿ classes. Utilizing this combined approach, the classification achieved an accuracy of 88.96%. The findings illuminate the dynamics of hate speech during significant events and underscore the imperative to analyze tweets to combat hate speech on social platforms, thereby fostering an inclusive and respectful online discourse. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.