NTU End-to-End Chinese Discourse Parser

Introduction

This page introduces an end-to-end Chinese discourse parser. The parser based on a unified framework by using recursive neural network (RvNN) to jointly model the subtasks including elementary discourse unit (EDU) segmentation, tree structure construction, center labeling, and sense labeling. To the best of our knowledge, this is the first open source toolkit for Chinese discourse parsing. We release the source code with a pre-trained model for the NLP community. The standalone toolkit can be integrated into subsequent applications without the need of external resources such as syntactic parser.

Usage

Follow the manual provided with the github package.

Example

The input is a paragraph in Simplified Chinese. For example:

• 去年，甘肃省国有大中型企业效益开始回升，一至十月，净亏损比上年同期减少五千多万元。全年全省完成工业增加值二百八十七点一亿元，比上年增长百分之十点三。与此同时，固定资产投资力度加大，投资总额达二百六十亿元，同比增长百分之二十一。

The output contains the EDU numbering and the discourse parsing tree constructing from the EDUs, associated with corresponding relation sense and centering:

Download

Click here to the github of the toolkit.