Factual Error Correction for Abstractive Summarization Models
This addresses the issue of factual inconsistency in abstractive summarization for users relying on accurate summaries, but it is incremental as it builds on existing error correction methods.
The paper tackles the problem of factual errors in summaries generated by neural abstractive summarization models by proposing a post-editing corrector module, which outperforms previous models on factual consistency evaluation on the CNN/DailyMail dataset.
Neural abstractive summarization systems have achieved promising progress, thanks to the availability of large-scale datasets and models pre-trained with self-supervised methods. However, ensuring the factual consistency of the generated summaries for abstractive summarization systems is a challenge. We propose a post-editing corrector module to address this issue by identifying and correcting factual errors in generated summaries. The neural corrector model is pre-trained on artificial examples that are created by applying a series of heuristic transformations on reference summaries. These transformations are inspired by an error analysis of state-of-the-art summarization model outputs. Experimental results show that our model is able to correct factual errors in summaries generated by other neural summarization models and outperforms previous models on factual consistency evaluation on the CNN/DailyMail dataset. We also find that transferring from artificial error correction to downstream settings is still very challenging.