Style versus Content: A distinction without a (learnable) difference?

Somayeh Jafaritazehjani (1, 2) , Gwénolé Lecorvé (1) , Damien Lolive (1) , John D Kelleher (2)

1 EXPRESSION - Expressiveness in Human Centered Data/Media
2 TU - Technological University [Dublin]

Textual style transfer involves modifying the style of a text while preserving its content. This assumes that it is possible to separate style from content. This paper investigates whether this separation is possible. We use sentiment transfer as our case study for style transfer analysis. Our experimental methodology frames style transfer as a multi-objective problem, balancing style shift with content preservation and fluency. Due to the lack of parallel data for style transfer we employ a variety of adversarial encoder-decoder networks in our experiments. Also, we use a probing methodology to analyse how these models encode style-related features in their latent spaces. The results of our experiments which are further confirmed by a human evaluation reveal an inherent trade-off between the multiple style transfer objectives and indicate that style cannot be usefully separated from content within these style-transfer systems.

Domaines: Intelligence artificielle [cs.AI]

Somayeh Jafaritazehjani, Gwénolé Lecorvé, Damien Lolive, John D Kelleher. Style versus Content: A distinction without a (learnable) difference?. International Conference on Computational Linguistics, Dec 2020, Virtual, Spain. ⟨hal-03112354⟩