Project by Sabrina Fonseca Pereira, Ida Maria Zachariassen, Magnus Sverdrup, Rasmus Bondo Hansen and Ruben Oliver Jonsman
This project explores how machine learning models, trained on Twitter data, perform when trying to classify the writer’s intention. The focus was split into two categories, one predicting irony and the other predicting stance regarding atheism.
We tested different machine learning models and developed different tokenizers to find out what combination yielded the best performance when classifying tweets.