English

Summary Refinement through Denoising

Computation and Language 2019-07-26 v1

Abstract

We propose a simple method for post-processing the outputs of a text summarization system in order to refine its overall quality. Our approach is to train text-to-text rewriting models to correct information redundancy errors that may arise during summarization. We train on synthetically generated noisy summaries, testing three different types of noise that introduce out-of-context information within each summary. When applied on top of extractive and abstractive summarization baselines, our summary denoising models yield metric improvements while reducing redundancy.

Keywords

Cite

@article{arxiv.1907.10873,
  title  = {Summary Refinement through Denoising},
  author = {Nikola I. Nikolov and Alessandro Calmanovici and Richard H. R. Hahnloser},
  journal= {arXiv preprint arXiv:1907.10873},
  year   = {2019}
}

Comments

RANLP 2019

R2 v1 2026-06-23T10:30:19.970Z