ABSTRACT
Hate speech is a problem that has become especially visible with the development of the Internet, including social media. Leaving it unresolved results in social and economic consequences, hence the attempts to limit this phenomenon. One of the activities supporting such processes may be automatic hate speech detection using text mining methods to block aggressive statements.
The aim of the chapter is to demonstrate and confirm the thesis about the effectiveness of text mining in the automatic detection of hate speech on the Internet. Moreover, the purpose is to compare different methods in terms of their effectiveness. The following text mining methods were compared: artificial neural network, naive Bayes classifier and support vector machine.
Considerations do not allow for the full automation of the process of eliminating hate speech, but it can significantly accelerate the moderation of the content on the Internet. The use of the proposed model in practice may bring several social (e.g. reduction of discrimination of minorities, social unrest) and economic (e.g. reduction of disturbances in the labor market and free movement of capital) benefits.
