Planning with Learned Entity Prompts for Abstractive Summarization
By Shashi Narayan, Yao Zhao, Joshua Maynez, Gonçalo Simões, Vitaly Nikolaev, Ryan T. McDonald. 2021
134 citations recorded by the source corpus.
Abstract We introduce a simple but flexible mechanism to learn an intermediate plan to ground the generation of abstractive summaries. Specifically, we prepend (or prompt) target summaries with entity chains—ordered sequences of entities mentioned in the summary. Transformer-based sequence-to-sequence models are then trained to generate the entity chain and then continue generating the summary conditioned on the entity chain and the input. We experimented with both pretraining and finetuning with this content planning objective. When evaluated on CNN/DailyMail, XSum, SAMSum, and BillSum, we demonstrate empirically that the grounded generation with the planning objective improves entity specificity and planning in summaries for all datasets, and achieves state-of-the-art performance on XSum and SAMSum in terms of rouge. Moreover, we demonstrate empirically that planning with entity chains provides a mechanism to control hallucinations in abstractive summaries. By prompting the decoder with a modified content plan that drops hallucinated entities, we outperform state-of-the-art approaches for faithfulness when evaluated automatically and by humans.
Subfields: Artificial Intelligence, Natural Language Processing
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Sebastian Riedel
- Hinrich Schütze
- Timo Schick
- Luke S. Zettlemoyer
- Hannaneh Hajishirzi
- Patrick Lewis
- Fabio Petroni
- Graham Neubig
In Natural Language Processing
- Sebastian Riedel
- Patrick Lewis
- Fabio Petroni
- Timo Schick
- Gautier Izacard
- Edouard Grave
- Hannaneh Hajishirzi
- Elizabeth Clark
Citations in the corpus
- Repairing the Cracked Foundation: A Survey of Obstacles in Evaluation Practices for Generated Text (2022, background): “While separate reasoning steps are starting to be incorporated into current NLG approaches (e.g., Puduppully et al., 2019; Narayan et al., 2021), there is no consensus for what a planning stage should look like or how to evaluate it, and all current evaluation practices are focused on output forms.”
- PEER: A Collaborative Language Model (2022, background): “The idea of having a separate planning stage has also been explored for other text generation tasks inlcuding summarization (Narayan et al., 2021), data-to-text generation (Moryossef et al.”