Zuo Zhuan Ancient Chinese Dataset for WSD

Update
We update the accuracy, you can check it in our file.
We also tried other Ancient Chinese models, you can check it in our file.

Published in NAACL SRW 2022
Paper link: https://aclanthology.org/2022.naacl-srw.17/

Authors: Xiaomeng Pan, Hongfei Wang, Mamoru Komachi, Teruaki Oka

Here are codes and our dataset of Zuo Zhuan Ancient Chinese Dataset for WSD mentioned in our paper.
Simple words for introducing our work are below, if you are interested in our research, please refer to our paper, Thanks.

Resource

Here we use two books: Kangxi Dict and Zuo Zhuan.

Kangxi Dictionary

Compiled in 1716, free for copyright, containing explanations for almost all the characters of the dynasties before Qing Dynasty.

Zuo Zhuan

One of the most famous ancient books, free for copyright. We used Zuo Zhuan following the previous study, and the resource is from wikisource.

Method

We apply 1-NN approach and GuwenBERT for WSD.

1-NN approach

Here we follow the approach from this paper: https://aclanthology.org/P19-1569/

GuwenBERT

you can check the model in this page: https://github.com/Ethan-yt/guwen-models

Licence

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.