cvejoski / Neural-Dynamic-Focused-Topic-Model

Code accompanying the paper: "Neural dynamic focused topic model." Proceedings of the AAAI Conference on Artificial Intelligence. Vol. 37. No. 11. 2023..
MIT License
1 stars 2 forks source link

Neural Dynamic Focused Topic Model

Code accompanying the paper: "Neural dynamic focused topic model." Proceedings of the AAAI Conference on Artificial Intelligence. Vol. 37. No. 11. 2023.

Installation of the library

In order to set up the necessary environment:

  1. create an environment drf with the help of [conda],

    conda env create -f environment.yaml
  2. activate the new environment with

    conda activate drf
  3. install drf with:

    
    python setup.py install # or `develop`
    python setup.py install # or

Data Preprocessing and preparation

All the data used should be in the folder {project_location}/data. The preprocessing of the *NIPS, ACL and UN datasets is done using the script located in {project_location}/scripts/preprocessing/dataset_preprocessor.py.

Training of models

The training of the models used in the paper is done by using the training scripts located in the {project_location}/scripts/train/{acl|nips|un|nips_perrone} folder.

Evaluating of the trained models

For evaluation of the trained models we use the script {project_location}/scripts/evaluate/topic_models_evaluate_all.py