Leveraging Large Pretrained Models for WebNLG 2020

By Xintong Li, Aleksandre Maskharashvili, Symon Jory Stevens-Guille, Michael. White. 2020

Subfields: Natural Language Processing, Generative Models

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Natural Language Processing

1. Soumya Batra
2. Ankit Arun
3. Shashank Jain
4. Peyman Heidari
5. Dmytro Okhonko
6. Michael. White
7. Payel Das
8. Xilun Chen

In Generative Models

1. Xilun Chen
2. Pierre L. Dognin
3. Sonal Gupta
4. Igor Melnyk
5. Inkit Padhi

Citations in the corpus

• Structure-to-Text Generation with Self-Training, Acceptability Classifiers and Context-Conditioning for the GEM Shared Task (2021, methodology): “To format the input data, we followed Li et al. (2020) in separating subjects, predicates and objects with separators while replacing un-derscores with spaces and removing quotes; we also prepended the category with a separator.”
• UniK-QA: Unified Representations of Structured and Unstructured Knowledge for Open-Domain Question Answering (2020, methodology): “In particular, we experiment with a top-ranked KB-to-text model (Li et al., 2020b) from the WebNLG 2020 challenge (Castro Ferreira et al., 2020), which is based on T5-large.”
• ReGen: Reinforcement Learning for Text and Knowledge Base Generation using Pretrained Language Models (2021, methodology): “2 nd place OSU Neural NLG (Li et al., 2020), the closest to our approach in spirit, used T5 and mBART PLMs to fine-tune after special data preprocessing.”