The Natural Language Generation Pipeline, Neural Text Generation and Explainability - Archive ouverte HAL Access content directly
Conference Papers Year : 2020

The Natural Language Generation Pipeline, Neural Text Generation and Explainability

Abstract

End-to-end encoder-decoder approaches to data-to-text generation are often black boxes whose predictions are difficult to explain. Breaking up the end-to-end model into submodules is a natural way to address this problem. The traditional pre-neural Natural Language Generation (NLG) pipeline provides a framework for breaking up the end-to-end encoder-decoder. We survey recent papers that integrate traditional NLG sub-modules in neural approaches and analyse their explainability. Our survey is a first step towards building explainable neural NLG models.
Fichier principal
Vignette du fichier
Submission_Workshop_INLG2020___NL4XAI___8_Dec_2020.pdf (137.21 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-03046206 , version 1 (08-12-2020)

Identifiers

  • HAL Id : hal-03046206 , version 1

Cite

Juliette Faille, Albert Gatt, Claire Gardent. The Natural Language Generation Pipeline, Neural Text Generation and Explainability. 2nd Workshop on Interactive Natural Language Technology for Explainable Artificial Intelligence, Dec 2020, Dublin (online), Ireland. ⟨hal-03046206⟩
247 View
452 Download

Share

Gmail Mastodon Facebook X LinkedIn More