Controlling Linguistic Style Aspects in Neural Language Generation

This is a re-implementation project carried out during my undergraduate studies at the University of Heidelberg in the course "Introduction to Neural Networks" on the topic "Stylized language generation".

The aim of the project is mainly to reimplement the approach of Ficler et al. (2017, "Controlling Linguistic Style Aspects in Neural Language Generation") and test other parameters, such as different LSTM architectures and smaller vocabulary sizes, especially using another corpus to ascertain whether the proposed method is scalable enough and can be replicated in other settings.