Creative GANs for generating poems, lyrics, and metaphors
This addresses the challenge of producing original and coherent creative text for applications in arts and entertainment, though it appears incremental as it adapts an existing GAN framework to this domain.
The paper tackled the problem of generating creative text like poems, lyrics, and metaphors, where traditional maximum likelihood optimization leads to generic and repetitive outputs, and reported better performance using a Generative Adversarial Network framework on diverse datasets.
Generative models for text have substantially contributed to tasks like machine translation and language modeling, using maximum likelihood optimization (MLE). However, for creative text generation, where multiple outputs are possible and originality and uniqueness are encouraged, MLE falls short. Methods optimized for MLE lead to outputs that can be generic, repetitive and incoherent. In this work, we use a Generative Adversarial Network framework to alleviate this problem. We evaluate our framework on poetry, lyrics and metaphor datasets, each with widely different characteristics, and report better performance of our objective function over other generative models.