INSET: Sentence Infilling with INter-SEntential Transformer
Abstract
Missing sentence generation (or sentence infilling) fosters a wide range of applications in natural language generation, such as document auto-completion and meeting note expansion. This task asks the model to generate intermediate missing sentences that can syntactically and semantically bridge the surrounding context. Solving the sentence infilling task requires techniques in natural language processing ranging from understanding to discourse-level planning to generation. In this paper, we propose a framework to decouple the challenge and address these three aspects respectively, leveraging the power of existing large-scale pre-trained models such as BERT and GPT-2. We empirically demonstrate the effectiveness of our model in learning a sentence representation for generation and further generating a missing sentence that fits the context.
Cite
@article{arxiv.1911.03892,
title = {INSET: Sentence Infilling with INter-SEntential Transformer},
author = {Yichen Huang and Yizhe Zhang and Oussama Elachqar and Yu Cheng},
journal= {arXiv preprint arXiv:1911.03892},
year = {2020}
}
Comments
Y.H. and Y.Z. contributed equally to this work. v2: published version with updated results and references