Adversarial NLI for Factual Correctness in Text Summarisation Models
This work addresses the issue of factual errors in text summarization for NLP applications, but it is incremental as it builds on prior ranking methods.
The authors tackled the problem of factual correctness in abstract summarization by applying the Adversarial NLI dataset to train NLI models, achieving significantly higher accuracy in ranking summaries based on entailment probabilities.
We apply the Adversarial NLI dataset to train the NLI model and show that the model has the potential to enhance factual correctness in abstract summarization. We follow the work of Falke et al. (2019), which rank multiple generated summaries based on the entailment probabilities between an source document and summaries and select the summary that has the highest entailment probability. The authors' earlier study concluded that current NLI models are not sufficiently accurate for the ranking task. We show that the Transformer models fine-tuned on the new dataset achieve significantly higher accuracy and have the potential of selecting a coherent summary.