XAlign: Cross-lingual Fact-to-Text Alignment and Generation for Low-Resource Languages
This addresses the need for automated text generation in low-resource languages for applications like Wikipedia, representing a novel approach as no prior work exists on cross-lingual alignment or generation for such languages.
The paper tackles the problem of generating descriptive text in low-resource languages from English fact triples, proposing unsupervised methods for cross-lingual alignment and introducing the XALIGN dataset with 0.45M pairs across 8 languages, including 5402 manually annotated pairs, along with baseline generation models.
Multiple critical scenarios (like Wikipedia text generation given English Infoboxes) need automated generation of descriptive text in low resource (LR) languages from English fact triples. Previous work has focused on English fact-to-text (F2T) generation. To the best of our knowledge, there has been no previous attempt on cross-lingual alignment or generation for LR languages. Building an effective cross-lingual F2T (XF2T) system requires alignment between English structured facts and LR sentences. We propose two unsupervised methods for cross-lingual alignment. We contribute XALIGN, an XF2T dataset with 0.45M pairs across 8 languages, of which 5402 pairs have been manually annotated. We also train strong baseline XF2T generation models on the XAlign dataset.