Cheap Translation for Cross-Lingual Named Entity Recognition

Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 2536–2545 Copenhagen, Denmark, September 7–11, 2017. c⃝2017 Association for Computational Linguistics Cheap Translation for Cross-Lingual Named Entity Recognition Stephen Mayhew Chen-Tse Tsai Dan Roth University of Illinois, Urbana-Champaign 201 N. Goodwin Urbana, Illinois, 61801 {mayhew2, ctsai12, danr}@illinois.edu Abstract Recent work in NLP has attempted to deal with low-resource languages but still as- sumed a resource level that is not present for most languages, e.g., the availabil- ity of Wikipedia in the target language. We propose a simple method for cross- lingual named entity recognition (NER) that works well in settings with very min- imal resources. Our approach makes use of a lexicon to “translate” annotated data available in one or several high resource language(s) into the target language, and learns a standard monolingual NER model there. Further, when Wikipedia is avail- able in the target language, our method can enhance Wikipedia based methods to yield state-of-the-art NER results; we evaluate on 7 diverse languages, improving the state-of-the-art by an average of 5.5% F1 points. With the minimal resources re- quired, this is an extremely portable cross- lingual NER approach, as illustrated using a truly low-resource language, Uyghur. 1 Introduction In recent years, interest in the natural language processing (NLP) community has expanded to in- clude multilingual applications. Although this uptick of interest has produced diverse annotated corpora, most languages are still classified as low- resource. In order to build NLP tools for low- resource languages, we either need to annotate data (a costly exercise, especially for languages with few native speakers), or find a way to use an- notated data in other languages in service to the cause. We refer to the latter techniques as cross- lingual techniques. In this paper, we address cross-lingual named German Spanish Dutch Avg Baseline 22.61 45.77 43.10 37.27 Previous SOA 48.12 60.55 61.56 56.74 Cheap Translation 57.53 65.18 66.50 62.65 Table 1: We show dramatic improvement on 3 Eu- ropean languages in a low-resource setting. More detailed results in Table 2 show that this improve- ment continues to a wide variety of languages. The baseline is a simple direct transfer model. The pre- vious state-of-the-art (SOA) is Tsai et al. (2016) entity recognition (NER). Prior methods (de- scribed in detail in Section 2) depend heavily on limited and expensive resources such as Wikipedia or large parallel text. Concretely, there are about 3800 written languages in the world.1 Wikipedia exists in about 280 languages, but most versions are too sparse to be useful. Parallel text may be found on an ad-hoc basis for some languages, but it is hardly a general solution. Religious texts, such as the Bible and the Koran, exist in many lan- guages, but the unique domain makes them hard to use. This leaves the vast majority of the world’s languages with no general method for NER. We propose a simple solution that requires only minimal resources. We translate annotated data in a high-resource language into a low-resource lan- …

Content truncated to 500 words