Better Translation + Split and Generate for Multilingual RDF-to-Text (WebNLG 2023)

Abstract

This paper presents system descriptions of our submitted outputs for WebNLG Challenge 2023. We use mT5 in multi-task and multilingual settings to generate more fluent and reliable verbalizations of the given RDF triples.

Furthermore, we introduce a partial decoding technique to produce more elaborate yet simplified outputs. Additionally, we demonstrate the significance of employing better translation systems in creating training data.

Keywords

• Translation
• Generate
• Multilingual
• RDF-to-Text
• WebNLG
• training
• data