Skip to content

zss019/CIMDL

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

15 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CIMDL

Introduction

Source code of paper: "Online Opinion Conflict Interaction Recognition Based on Dependent Multi-Task Deep Learning". Accepted by ICWSM 2026 (International AAAI Conference on Web and Social Media).

Part of code (particularly data processing and loading) is based on the [Multi-architecture Open-source Chinese Text Classification] (https://github.com/649453932/Chinese-Text-Classification-Pytorch) project and has undergone significant modifications and adjustments.

Dataset

ScienceNet_dataset and zhihu_dataset are two benchmark datasets we have constructed, which have been divided into train/dev/test sets.

We crawled users' opinion interaction records and manually annotated them.

Environments

  • Python 3.10

  • PyTorch 2.1.0

  • RTX 3090 GPU

  • CUDA 12.1

    Recommended to use GPU for training (such as Google Colab, Kaggle, AutoDL, Aliyun, etc.)

File Structure

The CIMDL.py file located in the models folder is the model file, which contains the definition scheme of the model.

The run.py, train_eval.py, and utils.py files are responsible for the project's operation, training, and processing respectively.

Additionally, the pre-trained model we used was downloaded from: https://huggingface.co/hfl/chinese-roberta-wwm-ext and https://huggingface.co/DMetaSoul/sbert-chinese-qmc-domain-v1.

Running

python run.py --model CIMDL

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages