Classical Tibetan corpus annotated for verb-argument dependency relations

This is a small hand-annotated partial treebank of Tibetan, primarily in CoNLL-U format. It builds upon the following corpus: Hill, Nathan W., & Garrett, Edward. (2017). A part-of-speech (POS) tagged corpus of Classical Tibetan [Data set]. Zenodo. http://doi.org/10.5281/zenodo.574878 This corpus differs from the above in three ways: The tagset has been converted from the SOAS tag system to the Universal Dependency part-of-speech tagset. We have added dependency relations between verbs and their argument. For some of the texts, English translations were available in digital form. These translations were manually aligned to the Tibetan texts and included in the CoNLL-U files. It was created as part of the AHRC-funded project Lexicography in Motion (PI Ulrich Pagel, 2017-2021).

本数据集为小型人工标注藏语局部树库（treebank），主要采用CoNLL-U格式。本数据集基于如下语料库构建：希尔（Nathan W. Hill）与加勒特（Edward）于2017年发布的《古典藏语词性标注（Part-of-Speech, POS）语料库》[数据集]，发布于Zenodo平台，DOI为10.5281/zenodo.574878。本数据集与上述语料库存在三处差异：其一，标记体系已从SOAS标记系统转换为通用依存（Universal Dependency）词性标记集；其二，新增了动词与其论元之间的依存关系；其三，部分文本配有数字版英文译文，我们将译文与藏语文本进行人工对齐后，一并纳入CoNLL-U格式文件中。本数据集是由AHRC资助的Lexicography in Motion（动态词典编纂）项目（项目负责人Ulrich Pagel，2017-2021）的研究成果之一。