Skip to content
Tag-enhanced Tree-Structured NN for discourse relation classification.
Branch: master
Clone or download
Latest commit 9b94421 Dec 6, 2017
Type Name Latest commit message Commit time
Failed to load latest commit information.
.gitignore open source Nov 24, 2017
LICENSE Create LICENSE Nov 24, 2017 update the readme Dec 6, 2017


This repo contains the code for a state-of-the-art discourse relation classifier, trained and tested on PDTB V2.

Here We implemented the Tree-LSTM and Tree-GRU models and enhanced them with pos-tag information. You can find more details in our paper at IJCNLP 2017: Tag-Enhanced Tree-Structured Neural Networks for Implicit Discourse Relation Classification.


We trained and and tested our models on PDTB V2 dataset. You should download them and convert the PDTB into a "pipe" delimited file using the built-in tool in the dataset.


To run the program, you can use the following command for preprocessing the data, prepare the formatted dataset and train the models:

python [--preprocess] [--prepare] [--train]

For detailed usage, run python -h to see the list of options.


MIT license

You can’t perform that action at this time.