Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/ucfnlp/summarization-dpp-capsnet


https://github.com/ucfnlp/summarization-dpp-capsnet

Last synced: about 1 month ago
JSON representation

Awesome Lists containing this project

README 

        
# Improving the Similarity Measure of Determinantal Point Processes for Extractive Multi-Document Summarization



We provide the source code for the paper **"[Improving the Similarity Measure of Determinantal Point Processes for Extractive Multi-Document Summarization](https://arxiv.org/abs/1906.00072)"**, accepted at ACL'19. If you find the code useful, please cite the following paper. 



    @inproceedings{cho-lebanoff-foroosh-liu:2019,

     Author = {Sangwoo Cho and Logan Lebanoff and Hassan Foroosh and Fei Liu},

     Title = {Improving the Similarity Measure of Determinantal Point Processes for Extractive Multi-Document Summarization},

     Booktitle = {Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (ACL)},

     Year = {2019}}



This repository contains the code for a similarity measure network using [Capsule network](https://github.com/XifengGuo/CapsNet-Keras).



## Dependencies

This code is developed with the following environment:

- [Python 2.7](https://www.anaconda.com/download/)

- [Keras 2.2.4](https://keras.io/)

- [Tensorflow 1.12.0 backend](https://www.tensorflow.org/install)

- `pip install -r requirements.txt`



## Train and evaluate on the CNN/DM summary pair dataset

### Set up directory for training/testing data

```

$ git clone https://github.com/sangwoo3/summarization-dpp-capsnet.git & cd summarization-dpp-capsnet

$ mkdir data & cd data

```



### Download the data

1. Download CNN/DM summary pair dataset from [HERE](https://drive.google.com/file/d/1GyvIHzoTat0xSZsdYJjocLji4eNnQ6NW/view?usp=sharing) and extract it under `/data` directory 

   * This summary dataset is pre-processed with 50k prevailing vocabularies in CNN/DM summary pair dataset. The label is 1 for a positive pair sentence, and 0 for a negative pair. The positive pair is a pair of a summary sentence and its most similar sentence in the source document that leads to the largest Rouge scores. The negative pair is a pair of the same summary sentence and a random sentence in the same document.

2. Download Glove word vectors of 50k vocabulary from [HERE](https://drive.google.com/file/d/1IpVT7LQ73_yqPYHaHHIAJ2iWPva0x2Uv/view?usp=sharing) and place it under `/data` directory

   * 6B tokens, 300d  Glove word vectors are used [LINK](http://nlp.stanford.edu/data/glove.6B.zip)

3. If you want raw CNN/DM summary dataset, download from [HERE](https://drive.google.com/file/d/1_c4AqnEct0HMg0VOWqupcO0_ijn-fJb0/view?usp=sharing). 

   * This data contains candiate summary sentences for each document. The data is pre-processed with the `preprocess.py` file to generate the above CNN/DM summary pair dataset.)



### Training

`$ python main_Capsnet.py`



### Testing

`$ python main_Capsnet.py --testing`



### Testing on STS dataset

`$ python main_Capsnet.py --testing --test_mode STS`



### Pre-trained Model

- Download the pre-trained model from [HERE](https://drive.google.com/file/d/11-Bc_BhBFRDTUWhM3ihiCm385GGkGWzZ/view?usp=sharing) and place it under `/result/capnet_sim` directory 

  - `/result/capnet_sim` is a default directory for training results

- Download the model fine-tuned on STS dataset from [HERE](https://drive.google.com/file/d/1LR6MncA2ViNVQqcMxxFR6fAYMUaoA6Rs/view?usp=sharing) 

  - This model is trained on CNN/DM summary pair dataset and then fine-tuned on STS.

  - It can be used to evaluate STS prediction accuracy.



### System summary

We provide our best system summaries of DUC04 and TAC11. They are generated with DPP and in the `system_summary` directory.

For DPP and multi-document dataset, we do not provide the code and dataset due to license. Please refer to [DPP](https://www.alexkulesza.com/) code and download [DUC 03/04](https://duc.nist.gov/) and [TAC 08/09/10/11](https://tac.nist.gov/data/index.html) dataset with your request and approval.



## License

This project is licensed under the BSD License - see the [LICENSE.md](LICENSE.md) file for details.