Data-to-Text Generation with Style Imitation

Abstract

Recent neural approaches to data-to-text generation have mostly focused on improving content fidelity while lacking explicit control over writing styles (e.g., word choices, sentence structures). More traditional systems use templates to determine the realization of text. Yet manual or automatic construction of high-quality templates is difficult, and a template acting as hard constraints could harm content fidelity when it does not match the record perfectly. We study a new way of stylistic control by using existing sentences as soft templates. That is, the model learns to imitate the writing style of any given exemplar sentence, with automatic adaptions to faithfully describe the content record. The problem is challenging due to the lack of parallel data. We develop a neural approach that includes a hybrid attention-copy mechanism, learns with weak supervisions, and is enhanced with a new content coverage constraint. We conduct experiments in restaurants and sports domains. Results show our approach achieves stronger performance than a range of comparison methods. Our approach balances well between content fidelity and style control given exemplars that match the records to varying degrees.

arXiv commentAccepted by EMNLP 2020 Findings. Significant updates over the previous version. Code & data are available at https://github.com/ha-lins/DTG-SI

References in corpus (13)

• Adam: A Method for Stochastic Optimization
• Sequence to Sequence Learning with Neural Networks
• Style Transfer from Non-Parallel Text by Cross-Alignment
• Harnessing Deep Neural Networks with Logic Rules
• Evaluating the State-of-the-Art of End-to-End Natural Language Generation: The E2E NLG Challenge
• Modeling Coverage for Neural Machine Translation
• A Retrieve-and-Edit Framework for Predicting Structured Outputs
• Content preserving text generation with attribute controls
• Delete, Retrieve, Generate: A Simple Approach to Sentiment and Style Transfer
• Learning to Select, Track, and Generate for Data-to-Text
• Retrieve and Refine: Improved Sequence Generation Models For Dialogue
• Shaping the Narrative Arc: An Information-Theoretic Approach to Collaborative Dialogue
• Reference-Aware Language Models

Cited by in corpus (5)

• A Survey of Knowledge-Enhanced Text Generation
• Texar: A Modularized, Versatile, and Extensible Toolkit for Text Generation
• Controllable Paraphrase Generation with a Syntactic Exemplar
• Select and Attend: Towards Controllable Content Selection in Text Generation
• Learning to Select Bi-Aspect Information for Document-Scale Text Content Manipulation