The field of online discourse analysis is rapidly evolving, with a growing focus on detecting and mitigating the effects of manipulation and misinformation. Recent research has highlighted the importance of considering the complexities of online communities and the various factors that influence collective minds. Innovations in machine learning and natural language processing are enabling the development of more sophisticated methods for analyzing and modeling online discourse. Notably, the use of multi-view autoencoders and other advanced techniques is improving the accuracy of fake news detection and hate speech classification. Furthermore, the creation of new datasets and frameworks, such as those focused on aporophobia and information disorder, is providing valuable insights into the dynamics of online discourse and the impacts of manipulation.

Noteworthy papers in this area include: The paper on multi-view autoencoders for fake news detection, which achieved significant improvements in classification performance. The paper on the dynamics of collective minds in online communities, which developed a computational model to describe and experiment with various influences on collective minds.