Balancing Lexical and Semantic Quality in Abstractive Summarization
This work addresses the gap between lexical and semantic metrics in summarization for NLP researchers, though it is incremental as it builds on existing re-ranking approaches.
The paper tackles the problem of exposure bias in abstractive summarization by proposing a re-ranking method that balances lexical and semantic quality, achieving a new state-of-the-art BERTScore of 89.67 on the CNN/DailyMail dataset.
An important problem of the sequence-to-sequence neural models widely used in abstractive summarization is exposure bias. To alleviate this problem, re-ranking systems have been applied in recent years. Despite some performance improvements, this approach remains underexplored. Previous works have mostly specified the rank through the ROUGE score and aligned candidate summaries, but there can be quite a large gap between the lexical overlap metric and semantic similarity. In this paper, we propose a novel training method in which a re-ranker balances the lexical and semantic quality. We further newly define false positives in ranking and present a strategy to reduce their influence. Experiments on the CNN/DailyMail and XSum datasets show that our method can estimate the meaning of summaries without seriously degrading the lexical aspect. More specifically, it achieves an 89.67 BERTScore on the CNN/DailyMail dataset, reaching new state-of-the-art performance. Our code is publicly available at https://github.com/jeewoo1025/BalSum.