Handling Rare Items in Data-to-Text Generation - Archive ouverte HAL Access content directly
Conference Papers Year : 2018

Handling Rare Items in Data-to-Text Generation

Abstract

Neural approaches to data-to-text generation generally handle rare input items using either delexicalisation or a copy mechanism. We investigate the relative impact of these two methods on two datasets (E2E and WebNLG) and using two evaluation settings. We show (i) that rare items strongly impact performance; (ii) that combining delexicalisation and copying yields the strongest improvement; (iii) that copying underperforms for rare and unseen items and (iv) that the impact of these two mechanisms greatly varies depending on how the dataset is constructed and on how it is split into train, dev and test.

Dates and versions

hal-02460010 , version 1 (29-01-2020)

Identifiers

Cite

Anastasia Shimorina, Claire Gardent. Handling Rare Items in Data-to-Text Generation. Proceedings of the 11th International Conference on Natural Language Generation, Nov 2018, Tilburg University, Netherlands. pp.360-370, ⟨10.18653/v1/W18-6543⟩. ⟨hal-02460010⟩
26 View
0 Download

Altmetric

Share

Gmail Mastodon Facebook X LinkedIn More