Interactions between politicians and public in social media

Keywords: Twitter, Natural Language Processing, Deep Learning, Graph theory
Project partners: Stefanie Bailer, Marius Sältzer
CeDA collaborator: Sébastien Boyer

• Characterizing the different flavors of interactions (positive, negative, offensive, etc...) between politicians and the general public on social media.

• Multilingual context dependent word embedding for sentences classification.
• Graph representation of social media interaction.

• Transformer-based Deep Neural Networks: Multilingual BERT or XML-RoBERTa.
• Graph Convolutional Neural Networks.

Figure 1. Tweets sentiment analysis based on context dependent word embedding, allows us to draw with confidence, directed edges representing emotional flow between politician tweets. ¶

Papers (see the Publications page for more details):

• Misogyny, politics, and social media determinants of hostile engagement against women parliamentarians on Twitter