Permalink
Cannot retrieve contributors at this time
Join GitHub today
GitHub is home to over 28 million developers working together to host and review code, manage projects, and build software together.
Sign up
Fetching contributors…
Cannot retrieve contributors at this time
| # UD_Japanese-KTC | |
| This treebank is produced by automatically converting Kaede treebank, | |
| which is built over the source material of Kyoto corpus. | |
| http://nlp.ist.i.kyoto-u.ac.jp/EN/index.php?Kyoto%20University%20Text%20Corpus | |
| ## Recovering data | |
| The data is provided in the CoNLL format, but original texts are | |
| stripped off due to the license issue. In order to recover the data | |
| with original texts, you need the corpus of Mainichi Shinbun 1995 (the | |
| same set of data for Kyoto Corpus). The Mainichi Shinbun data can be | |
| obtained from Nichigai Associates: | |
| http://www.nichigai.co.jp/sales/mainichi/mainichi-data.html | |
| The corpus is obtained by running the following command: | |
| ./autoconv -d MAINICHI_DIR | |
| where `MAINICHI_DIR` denotes the directory of the files of Mainichi | |
| Shinbun 1995. In order to run this command, you need Perl and | |
| Python. | |
| The program for extracting texts from Mainichi Shinbun 1995 is | |
| borrowed from the Kyoto Corpus project. | |
| ## Spliting | |
| Each data set contains UD annotations for the following sections in Kyoto Corpus. | |
| training: articles from January 1st to 4th and editorials from January to February | |
| development: articles on January 6th and editorials in April | |
| test: articles on January 5th and editorials in March | |
| ## Citation | |
| You are encouraged to cite the following paper when you refer to the | |
| Universal Dependencies Japanese Treebank. | |
| Hiroshi Kanayama, Yusuke Miyao, Takaaki Tanaka, Shinsuke Mori, | |
| Masayuki Asahara, and Sumire Uematsu. A Draft Proposal for Universal | |
| Dependencies Japanese Treebank. In the Proceedings of the 21st annual | |
| meeting for Gengo Shori Gakkai (The Association for Natural Language | |
| Processing). 2015. (In Japanese) | |
| ## Changelog | |
| * 2015-11-15 v1.2 | |
| * First release in UD | |
| === Machine-readable metadata ================================================= | |
| Data available since: UD v1.2 | |
| License: CC BY-SA 4.0 | |
| Includes text: no | |
| Genre: news | |
| Lemmas: converted from manual | |
| UPOS: converted from manual | |
| XPOS: not available | |
| Features: converted from manual | |
| Relations: converted from manual | |
| Contributors: Asahara, Masayuki; Kanayama, Hiroshi; Matsumoto, Yuji; Miyao, Yusuke; Mori, Shunsuke; Tanaka, Takaaki; Uematsu, Sumire | |
| Contributing: elsewhere | |
| Contact: hkana@jp.ibm.com | |
| =============================================================================== |