GANs have been shown to perform exceedingly well on tasks pertaining to image generation and style transfer. In the field of language modelling, word embeddings such as GLoVe and word2vec are state-of-the-art methods for applying neural network models on textual data. Attempts have been made to utilize GANs with word embeddings for text generation. We present an approach to text generation using Skip-Thought sentence embeddings with GANs based on gradient penalty functions and f-measures. The proposed architecture aims to reproduce writing style in the generated text by modelling the way of expression at a sentence level across all works of an author. We ran extensive experiments in different embedding settings on a variety of tasks including conditional text generation and language generation. Our model outperforms baseline text generation networks across several automated evaluation metrics like BLEU-n, METEOR and ROUGE. We also demonstrate wide applicability and effectiveness in real life tasks through human judgement scores.
If you found this work to be useful, please cite it as: