Summary Refinement through Denoising
Nikola I. Nikolov | Alessandro Calmanovici | Richard Hahnloser
Proceedings of the International Conference on Recent Advances in Natural Language Processing (RANLP 2019)
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.