Introducing the Asian Language Treebank (ALT)

Introducing the Asian Language Treebank (ALT) Ye Kyaw Thu†, Win Pa Pa‡, Masao Utiyama†, Andrew Finch†, Eiichiro Sumita† †Advanced Speech Translation Research and Development Promotion Center, NICT, Kyoto, Japan ‡Natural Language Processing Lab, University of Computer Studies, Yangon, Myanmar, {yekyawthu, multiyama, andrew.finch, eiichiro.sumita}@nict.go.jp, winpapa@ucsy.edu.mm Abstract This paper introduces the ALT project initiated by the Advanced Speech Translation Research and Development Promotion Center (ASTREC), NICT, Kyoto, Japan. The aim of this project is to accelerate NLP research for Asian languages such as Indonesian, Japanese, Khmer, Laos, Malay, Myanmar, Philippine, Thai and Vietnamese. The original resource for this project was English articles that were randomly selected from Wikinews. The project has so far created a corpus for Myanmar and will extend in scope to include other languages in the near future. A 20000-sentence corpus of Myanmar that has been manually translated from an English corpus has been word segmented, word aligned, part-of-speech tagged and constituency parsed by human annotators. In this paper, we present the implementation steps for creating the treebank in detail, including a description of the ALT web-based treebanking tool. Moreover, we report statistics on the annotation quality of the Myanmar treebank created so far. Keywords: Asia Language Tree Bank (ALT), Myanmar language, word segmentation, alignment, part-of-speech tagging, tree 1. Introduction We introduce and describe ongoing work in the cre- ation of the Asian Language Treebank (ALT) corpus. Although the ALT corpus is mainly designed for statis- tical machine translation (SMT) of Asian languages, it will be useful in general for natural language processing (NLP) research. The corpus contains essential infor- mation for NLP tasks such as word segmentation, word alignment to parallel English, part-of-speech (POS) tagging and constituency parse trees. Additionally, this will be the first open Asian language treebank corpus. The ALT project is one of the language re- source development projects of ASTREC and aims to accelerate research of NLP for Asian languages. Cur- rently there is no publicly available POS-tagged and constituency tree corpus for most of the Asian lan- guages and thus the APT project was created. In this paper, we present the ALT corpus builing for Myanmar (the official language of the Republic of the Union of Myanmar) in detail. We will also introduce the ALT web-based software tool that was developed for annotating manual translations with word segmen- tation, word alignment, POS tagging and parse trees. This paper is organized as follows: Section 2. gives an overview of the design of the ALT corpus. In Section 3. the rules and development steps for the Myanmar lan- guage are described in detail. In Section 4., we give a brief introduction to the Asian Language Treebank- ing Tool (ALT Tool). In Section 5. an evaluation of the quality of the work in progress is given. Finally, Section 6. concludes and describes how we expect the ALT project to progress in the near future. 2. Overview of the Asian Language Treebank (ALT) ASTREC, plans to coordinate the development of the ALT Corpus between 2014 to 2018. …

Content truncated to 500 words