MAGNeto

MAGNeto: An Efficient Deep Learning Method for the Extractive Tags Summarization Problem

Downloading NUS-WIDE dataset

• Official: https://lms.comp.nus.edu.sg/wp-content/uploads/2019/research/nuswide/NUS-WIDE.html
• Unofficial: http://cs-people.bu.edu/hekun/data/TALR/NUSWIDE.zip (recommended for downloading all images)

Data preparation

Moving images to a single directory

./data/nus_wide/notebooks/Move\ Images.ipynb

Preparing tag data

./data/nus_wide/notebooks/Prepare\ Tag\ Data.ipynb

Setting up the environment

pip install -U pip
pip install -r requirements.txt

Generating label for raw data

• Step 1: Reconfigure scripts/start_preprocess.sh

  To list all configurable parameters, run

  python preprocess.py -h

• Step 2: Run

  bash scripts/start_preprocess.sh

Training the model

• Step 1: Reconfigure scripts/start_train.sh

  To list all configurable parameters, run

  python train.py -h

• Step 2: Run

  bash scripts/start_train.sh

Inferring test data

• Step 1: Reconfigure scripts/start_infer.sh

  To list all configurable parameters, run

  python infer.py -h

• Step 2: Run

  bash scripts/start_infer.sh

Reference

Please acknowledge the following paper in case of using this code as part of any published research:

"MAGNeto: An Efficient Deep Learning Method for the Extractive Tags Summarization Problem." Hieu Trong Phung, Anh Tuan Vu, Tung Dinh Nguyen, Lam Thanh Do, Giang Nam Ngo, Trung Thanh Tran, Ngoc C. Lê.

@article{Hieu2020,
    title={MAGNeto: An Efficient Deep Learning Method for the Extractive Tags Summarization Problem},
    author={Hieu Trong Phung and Anh Tuan Vu and Tung Dinh Nguyen and Lam Thanh Do and Giang Nam Ngo and Trung Thanh Tran and Ngoc C. L\^{e}},
    journal={arXiv preprint arXiv:2011.04349},
    year={2020}
} 

License

The code is released under the GPLv3 License.