Generating Text from Uniform Meaning Representation

TL;DR

This paper explores methods for generating text from multilingual Uniform Meaning Representation (UMR) graphs, leveraging AMR technologies and fine-tuning models, achieving promising multilingual results despite limited UMR data.

Contribution

It introduces the first approaches to UMR-to-text generation, including a baseline, a conversion pipeline, and fine-tuning strategies using existing AMR technologies.

Findings

Achieved BERTscore of 0.825 for English

Achieved BERTscore of 0.882 for Chinese

Fine-tuning improves UMR-to-text generation effectiveness

Abstract

Uniform Meaning Representation (UMR) is a recently developed graph-based semantic representation, which expands on Abstract Meaning Representation (AMR) in a number of ways, in particular through the inclusion of document-level information and multilingual flexibility. In order to effectively adopt and leverage UMR for downstream tasks, efforts must be placed toward developing a UMR technological ecosystem. Though only a small amount of UMR annotations have been produced to date, in this work, we investigate the first approaches to producing text from multilingual UMR graphs. Exploiting the structural similarity between UMR and AMR graphs and the wide availability of AMR technologies, we introduce (1) a baseline approach which passes UMR graphs to AMR-to-text generation models, (2) a pipeline conversion of UMR to AMR, then using AMR-to-text generation models, and (3) a fine-tuning approach for both foundation models and AMR-to-text generation models with UMR data. Our best performing models achieve multilingual BERTscores of 0.825 for English and 0.882 for Chinese, a promising indication of the effectiveness of fine-tuning approaches for UMR-to-text generation even with limited UMR data.