Deep Learning as a Digital Tool for the Detection and Prevention of Cyberbullying

The present research aimed to explore the state-of-the-art of deep learning (DL) techniques focused on the automatic detection of offensive behavior on the Internet, a problematic also known as cyberbullying. For this purpose, the most important DL algorithms were analyzed, as well as the studies conducted with them in order to combat cyberbullying. The information presented in this chapter was collected from a bibliographic search through indexed journals and databases such as Scopus, using keywords that included “cyberbullying”, “Deep learning”, “applications” in order to find relevant scientific articles on this topic. As a result of this analysis, it was concluded that DL presents a novel and effective alternative against cyberbullying. Given the current state of this technology, explored in this chapter, it can be stated that it will become an important tool in the future to successfully combat this problem, although there are still some aspects to be improved.