Applying Machine Learning Models to Classify Xenophobic Tweets Against Asians, With Data Analysis of Hate Crimes

This paper offers insight to the COVID-19 pandemic and its effect on people's attitudes towards certain minority groups, particularly Asians, Asian-Americans, and Pacific Islanders. With the Coronavirus first being identified in Wuhan, China, xenophobia, and racism towards groups pertaining to the supposed origins of the COVID-19 pandemic have been on the rise. Along with the violent physical attacks on these groups, this paper will focus on the online hate and xenophobia that Asians face due to their race, ethnicity, country of origin, and/or others. In this paper, Python is employed as the primary programming language; external libraries such as pandas, NumPy, sklearn, WordCloud, and matplotlib are imported for handling data. In analyzing the racism against Asians, keywords such as “Asian Hate,” “Hate Crime” and “anti-Asian” are utilized, and the Python programming language is employed to sift through Google News articles with these keywords and identify patterns in the words’ usages. Furthermore, the frequencies of the keywords’ usages on online platforms such as Twitter are also analyzed in the form of comma-separated files, with patterns of usage over time before and after the COVID-19 pandemic began being identified. Randomly selected tweets are classified into five categories: anti-Asian, not anti-Asian, not English, hate against others racial groups, and support towards Asians. These tweets are classified by artificial intelligence using machine learning methods of logistic regression, support vector machine, and Naive Bayes; the artificial intelligence was taught using pre-classified data sets. Classified tweets represent the implication and relevance between the tweets and xenophobia. This classification model of xenophobia is expected to be used in social media content censoring and enhance the internet chatting etiquette. The goal of this classification model is to terminate anti-Asian hatred and lower the overall level of societal racism.