부산대학교 도서관

Hate Speech is one of the most dangerous problems which disturbs social and religious harmony. Social media which plays a vital role in our lives has now become a responsible element in spreading hate speech. This work presents a comprehensive analysis of hate speech on Twitter, using Tweepy, an open-source Twitter API, to fetch relevant data. Our study sought to investigate the behavioral patterns of users responsible for spreading hateful comments online. After collecting a sample of tweets, standard sentiment analysis techniques have been leveraged to segregate the tweets as per their positive and negative sentiments. Using Affin lexicon and TextBlob libraries, the polarity and subjectivity of the tweets have been identified. The k-means clustering algorithm is utilized to group the data into positive and negative clusters. These clusters have been considered as communities of the users and further analyzed. Standard metrics indicate a strong community formation in our proposed approach. The feasibility of clustering the features of user expressions toward community formation in social networks is explored in this work. Our findings provide valuable insights into the prevalence and nature of hate speech on social media and offer suggestions for addressing this critical issue.