Automatic identification of misogynistic sentiments on social networks

Abstract

The number of social media users is constantly growing. Automatic sentiment analysis in unstructured text using artificial intelligence is a tool that allows organizations to identify areas for improvement based on users' opinions. Natural language processing enables computational treatment of these opinions through emotion analysis and polarity identification in texts. This work focuses on the automatic identification of misogyny in unstructured texts using different classification scenarios and machine learning methods, as well as the use of meta classifiers, with the aim of identifying the pre-processing and processing techniques that lead to the best performance in this task. The results obtained show the effectiveness of automatic sentiment analysis tools on Twitter and its importance in better understanding complex social phenomena.