/NMTM

Code for Learning Multilingual Topics with Neural Variational Inference (NLPCC2020)

Primary LanguagePython

Check our latest topic modeling toolkit TopMost !

Code for Learning Multilingual Topics with Neural Variational Inference

NLPCC2020 paper

Usage

0. Download datasets

Google Drive

Unzip and move the datasets to ./data

1. Prepare environment

python==3.6
tensorflow-gpu==1.13.1
scipy==1.5.2
scikit-learn==0.23.2

2. Training

python run.py --data_dir data/{dataset} --output_dir output/{dataset}

3. Evaluation

  1. Topic coherence: topic_interpretability

  2. Topic diversity:

     python utils/TU.py --data_path {path of topic word file}

  3. Classification: use the SVM of scikit-learn.

Citation

If you want to use our code, please cite as

@inproceedings{Wu2020,
    author = {Wu, Xiaobao and Li, Chunping and Zhu, Yan and Miao, Yishu},
    booktitle = {International Conference on Natural Language Processing and Chinese Computing},
    title = {{Learning Multilingual Topics with Neural Variational Inference}},
    year = {2020}
}