Cross-Lingual Dependency Parsing with Late Decoding for Truly Low-Resource Languages
This work addresses dependency parsing for truly low-resource languages, representing an incremental advancement in cross-lingual NLP.
The paper tackled the problem of information loss in cross-lingual dependency parsing due to early decoding by introducing an end-to-end graph-based neural network that projects edge score matrices, resulting in a 2.25% absolute improvement averaged across 10 languages compared to previous state-of-the-art methods.
In cross-lingual dependency annotation projection, information is often lost during transfer because of early decoding. We present an end-to-end graph-based neural network dependency parser that can be trained to reproduce matrices of edge scores, which can be directly projected across word alignments. We show that our approach to cross-lingual dependency parsing is not only simpler, but also achieves an absolute improvement of 2.25% averaged across 10 languages compared to the previous state of the art.