Evaluation of NMT-Assisted Grammar Transfer for a Multi-Language Configurable Data-to-Text System
This addresses the problem of efficient multilingual text generation for domain-specific applications like sports reporting, though it is incremental as it builds on existing NMT and rule-based NLG methods.
The paper tackled multilingual data-to-text generation by using Neural Machine Translation (NMT) with human review to translate grammatical configurations, enabling scalable text generation without ongoing human intervention. Evaluation on the SportSett:Basketball dataset showed the system performs well, emphasizing grammatical correctness in translations.
One approach for multilingual data-to-text generation is to translate grammatical configurations upfront from the source language into each target language. These configurations are then used by a surface realizer and in document planning stages to generate output. In this paper, we describe a rule-based NLG implementation of this approach where the configuration is translated by Neural Machine Translation (NMT) combined with a one-time human review, and introduce a cross-language grammar dependency model to create a multilingual NLG system that generates text from the source data, scaling the generation phase without a human in the loop. Additionally, we introduce a method for human post-editing evaluation on the automatically translated text. Our evaluation on the SportSett:Basketball dataset shows that our NLG system performs well, underlining its grammatical correctness in translation tasks.