Comparison of Assorted Models for Transliteration

Abstract

We report the results of our experiments in the context of the NEWS 2018 Shared Task on Transliteration. We focus on the comparison of several diverse systems, including three neural MT models. A combination of discriminative, generative, and neural models obtains the best results on the development sets. We also put forward ideas for improving the shared task.

String Transduction with Target Language Models and Insertion Handling

“…Table 2 shows that DTLM outperforms the other systems by a large margin thanks to its ability to leverage a target word list. Additional results are reported by Najafi et al (2018b).…”

Combining Neural and Non-Neural Methods for Low-Resource Morphological Reinflection

“…http://www.speech.cs.cmu.edu/SLM/toolkit.html 9 DTLM was also succesfully used in the NEWS 2018 shared task on transliteration(Najafi et al, 2018b).…”

Cognate Projection for Low-Resource Inflection Generation

“…In the CoNLL-SIGMORPHON 2018 Shared Task on Universal Morphological Reinflection (Cotterell et al, 2018), DTLM was our best performing individual system. It was also successfully used in the NEWS 2018 shared task on transliteration (Najafi et al, 2018b).…”