aitsc / text-representation

Introduction

This code is the implementation of the paper "Coherence-Based Distributed Document Representation Learning for Scientific Documents".

Run

Installation environment

conda create -n ctpe python=3.6
conda activate ctpe
git clone https://github.com/aitsc/text-representation.git
cd text-representation
pip install https://download.pytorch.org/whl/cu90/torch-0.3.1-cp36-cp36m-linux_x86_64.whl
pip install allennlp==0.4.3
pip install -r requirements.txt

Download data

• Link: https://pan.baidu.com/s/1EEJk0_P55Ov5ReXsmyVZPA Password: rkh0
• Place the data folder in the code directory.

Test

• Modify the ap and model address in the code _av_CTE.py.
• python _av_CTE.py

Citation

If you find this code useful, please cite the following paper:

@article{tan2022coherence,
  title = {Coherence-Based Distributed Document Representation Learning for Scientific Documents},
  author = {Tan, Shicheng and Zhao, Shu and Zhang, Yanping},
  journal = {arXiv},
  year = {2022},
  type = {Journal Article}
}