Enriching the WebNLG corpus

By Thiago Castro Ferreira, Diego Moussallem, Emiel Krahmer, Sander Wubben. 2018

59 citations recorded by the source corpus.

This paper describes the enrichment of WebNLG corpus (Gardent et al., 2017a,b), with the aim to further extend its usefulness as a resource for evaluating common NLG tasks, including Discourse Ordering, Lexicalization and Referring Expression Generation. We also produce a silver-standard German translation of the corpus to enable the exploitation of NLG approaches to other languages than English. The enriched corpus is publicly available.

Subfields: Artificial Intelligence, Evaluation & Benchmarks, Natural Language Generation

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

In Evaluation & Benchmarks

In Natural Language Generation

Citations in the corpus

DART: Open-Domain Structured Data Record to Text Generation (2020, neutral): “Castro Ferreira et al. (2018) pointed out the bene-ﬁts of delexicalization for NLG models as it allows the model to describe categories instead of noisy entity names (e.g. "President" instead of "Barack Obama").”
Barack’s Wife Hillary: Using Knowledge Graphs for Fact-Aware Language Modeling (2019, background): “…much more aligned to the data-to-text task (Ahn et al., 2016; Lebret et al., 2016; Wiseman et al., 2017; Yang et al., 2017; Gardent et al., 2017; Ferreira et al., 2018), where a small table-sized KB is provided to generate a short piece of text; we are interested in language models that…”
Automatic Text Evaluation through the Lens of Wasserstein Barycenters (2021, background): “…task-oriented dialogue datasets (i.e., BAGEL (Mairesse et al., 2010), SFHOTEL (Wen et al., 2015)), we focus on the WebNLG challenge (PerezBeltrachini et al., 2016; Gardent et al., 2017; Ferreira et al., 2018, 2020) as sentence available in this challenge is more representative of progress of NLG4.”
Bridging the Structural Gap Between Encoding and Decoding for Data-To-Text Generation (2020, methodology): “While Step-By-Step uses heuristic string matching to extract plans from the referenced sentences, other methods (GRU and transformer), as well as ours, use plans provided in the enriched WebNLG dataset (Castro Ferreira et al., 2018).”
Neural data-to-text generation: A comparison between pipeline and end-to-end architectures (2019, methodology): “We used version 1.5 of the augmented WebNLG corpus (Castro Ferreira et al., 2018b) 4 to evaluate the steps of our pipeline approach.”
Decoder-Only or Encoder-Decoder? Interpreting Language Model as a Regularized Encoder-Decoder (2023, methodology): “We investigate the performance on three datasets, namely the WebNLG (Perez-Beltrachini et al., 2016; Gardent et al., 2017a,b; Castro Ferreira et al., 2018; Shimorina et al., 2019) dataset (CC BY-NC-SA 4.0 License), E2E (Novikova et al., 2017; Duˇsek et al., 2019) dataset, and the WITA (Fu et al.,…”
Plan-then-Generate: Controlled Data-to-Text Generation via Planning (2021, methodology): “In addition, we examine two ablated variants of our 4The human-generated plans are provided in the enriched WebNLG dataset (Ferreira et al., 2018). content planner by either removing the CRF layer (w/o CRF) or using randomly initialized parameters instead of the pre-trained BERT (w/o PLMs).”
Improving Text-to-Text Pre-trained Models for the Graph-to-Text Task (2020, methodology): “The KG-to-text data we use in this work is the enriched version of WebNLG dataset (Ferreira et al., 2018).”
Non-neural Models Matter: a Re-evaluation of Neural Referring Expression Generation Systems (2022, methodology): “The definition of each criterion was taken from Castro Ferreira et al. (2018a).”