abstract
- © 2021 ACM.Violence Against Women is a phenomenon that has grown in the last decades. In this research we address this phenomenon in Digital Space as Digital Violence Against Women. We trained a Support Vector Machine Classifier that predicted a 20.16% presence of Digital Violence Against Women in a Self-constructed Tweet dataset containing over 10 Million Spanish-language tweets in Mexico. A time series was constructed on the percentage of Digital Violence Against Women in tweets per day. Findings indicate increasing presence of violence on specific dates of fight for Women's Rights, such as 25th of November and 8th of March. Finally, forecasts under an ARIMA model gave a Root Mean Squared Error of 0.0062.