From Data Pre-processing to Hate Speech Detection: An Interdisciplinary Study on Women-targeted Online Abuse
Rutuja G. Rathod, Yashoda Barve, Jatinderkumar R. Saini, Sourav Rathod · 2023
The term “misogyny” conveys hatred and disrespect for women. It is a type of sexism that stems from the idea that women are less valuable than men and can take many forms, such as verbal abuse, sexual harassment, and physical violence. One area where misogyny is particularly prevalent is in online spaces. Women who speak out and share their experiences online often face vicious attacks, including violence, rape, and death threats. This phenomenon, often referred to as “online harassment,” can have devastating effects on women’s mental health and well-being. Despite the growing recognition of this issue, there is still a lack of understanding about the nature and extent of hate speech toward women on social media. In this paper, the authors aim to address this gap by analyzing a dataset of tweets on Twitter containing hate speech toward women. The authors introduced the Measuring Hate Speech corpus, a dataset used while studying hate speech towards women. From this dataset, the authors extract some of them to build the model. And authors proposed Machine Learning algorithms like Logistic Regression with 0.92% accuracy. And to look further to fill the gap, the authors developed Support Vector Machine (SVM) algorithm, Topic Modeling technique to extract the topics from the corpus. This study contributes to a better understanding of hate speech on social media platforms as Twitter informs strategies for resisting this pervasive problem.