Deciphering Related Languages

Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 2513–2518 Copenhagen, Denmark, September 7–11, 2017. c⃝2017 Association for Computational Linguistics Deciphering Related Languages Nima Pourdamghani, Kevin Knight Information Sciences Institute & Department of Computer Science University of Southern California {damghani,knight}@isi.edu Abstract We present a method for translating texts between close language pairs. The method does not require parallel data, and it does not require the languages to be written in the same script. We show results for six language pairs: Afrikaans/Dutch, Bosnian/Serbian, Danish/Swedish, Mace- donian/Bulgarian, Malaysian/Indonesian, and Polish/Belorussian. We report BLEU scores showing our method to outperform others that do not use parallel data. 1 Introduction Statistical Natural Language Processing (NLP) tools often need large amounts of training data in order to achieve good performance. This limits the use of current NLP tools to a few resource- rich languages. Assume an incident happens in an area with a low-resource language, known as the Incident Language (IL). For a quick response, we need to build NLP tools with available data, as finding or annotating new data is expensive and time consuming. For many languages this means that we only have a small amount of often out-of- domain parallel data (e.g. a Bible or Ubuntu man- ual), some monolingual data and almost no anno- tation such as part of speech tags. Fortunately, many low-resource languages have one or more higher-resource, closely Related Lan- guages (RL). Examples of such IL/RL pairs are Afrikaans/Dutch and Bosnian/Serbian. A natural idea is to use RL resources to improve the task for IL. But this requires some kind of conversion between RL and IL. Assume the required NLP ca- pability is named entity tagging. If we can con- vert RL to IL, we can convert all RL training data along with annotations into IL and train the tagger for IL. Or, if we can convert IL to RL we can use the potentially existing RL named entity tagger on converted IL data and project back the tags. Following this idea, Currey et al. (2016) use a rule-based translation system to convert Italian and Portuguese into Spanish, to improve Span- ish (here, IL) language modeling, Nakov and Ng (2009) convert RL/English parallel data to IL/English where both RL and IL have Latin or- thography to improve IL/English machine trans- lation. Hana et al. (2006) use cognates to adapt Spanish resources to Brazilian Portuguese to train a part-of-speech tagger. Mann and Yarowsky (2001) use Spanish/Portuguese cog- nates to convert an English/Spanish lexicon to En- glish/Portuguese. These works prove the useful- ness of RL data to improve NLP for IL, but they are designed for specific tasks and IL/RL pairs. In this paper we propose a universal method for translating texts between closely related lan- guages. We assume that IL and RL are mostly cognates, having roughly the same word order. Our method is orthography-agnostic for alphabetic systems, and crucially, it does not need any paral- lel data. From now on, we talk about converting RL …

Content truncated to 500 words