Zero-shot Sonnet Generation with Discourse-level Planning and Aesthetics Features
Yufei Tian, Nanyun Peng

TL;DR
This paper introduces a zero-shot sonnet generation framework that plans discourse, incorporates aesthetics, and enforces poetic constraints without requiring training on poetic data, resulting in more coherent and creative poems.
Contribution
A novel hierarchical, zero-shot framework for sonnet generation that combines discourse planning, aesthetic features, and constrained decoding without training on poetic corpora.
Findings
Outperforms baseline methods in coherence and creativity
Generates sonnets that meet poetic constraints effectively
Enhances poetic quality without training on poem datasets
Abstract
Poetry generation, and creative language generation in general, usually suffers from the lack of large training data. In this paper, we present a novel framework to generate sonnets that does not require training on poems. We design a hierarchical framework which plans the poem sketch before decoding. Specifically, a content planning module is trained on non-poetic texts to obtain discourse-level coherence; then a rhyme module generates rhyme words and a polishing module introduces imagery and similes for aesthetics purposes. Finally, we design a constrained decoding algorithm to impose the meter-and-rhyme constraint of the generated sonnets. Automatic and human evaluation show that our multi-stage approach without training on poem corpora generates more coherent, poetic, and creative sonnets than several strong baselines.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsArtificial Intelligence in Games · Topic Modeling · Multimodal Machine Learning Applications
