Moses and the Character-Based Random Babbling Baseline: CoAStaL at AmericasNLP 2021 Shared Task Article Swipe

PDF

Related Concepts

Babbling Baseline (sea) Character (mathematics) Task (project management) Computer science Speech recognition Artificial intelligence Natural language processing Linguistics Mathematics Philosophy Engineering Political science Systems engineering Law Geometry

Marcel Bollmann , Rahul Aralikatte , Héctor Murrieta Bello , Daniel Hershcovich , Miryam de Lhoneux , Anders Søgaard ·

YOU? · · 2021 · Open Access · · DOI: https://doi.org/10.18653/v1/2021.americasnlp-1.28 · OA: W3166307066

We evaluated a range of neural machine translation techniques developed specifically for low-resource scenarios. Unsuccessfully. In the end, we submitted two runs: (i) a standard phrase-based model, and (ii) a random babbling baseline using character trigrams. We found that it was surprisingly hard to beat (i), in spite of this model being, in theory, a bad fit for polysynthetic languages; and more interestingly, that (ii) was better than several of the submitted systems, highlighting how difficult low-resource machine translation for polysynthetic languages is.