Diversifying-Dialogue-Generation-with-Non-Conversational-Text

Implementation for the paper Diversifying Dialogue Generation with Non-Conversational Text on English

Adapt from

@inproceedings{su-etal-2020-diversifying,
    title = "Diversifying Dialogue Generation with Non-Conversational Text",
    author = "Su, Hui  and
      Shen, Xiaoyu  and
      Zhao, Sanqiang  and
      Xiao, Zhou  and
      Hu, Pengwei  and
      Zhong, Randy  and
      Niu, Cheng  and
      Zhou, Jie",
    booktitle = "Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics",
    month = jul,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/2020.acl-main.634",
    doi = "10.18653/v1/2020.acl-main.634",
    pages = "7087--7097"
}

Code Structure

|
|—— .gitignore
|—— README.md
|—— LICENSE
|—— scripts
|   |-- initialization.sh
|   |-- preprocess_initialization.sh
|   |-- run_classifier.sh
|   |—— run_forward.sh
|   |—— get_forward_data.sh
|—— src
|   |—— forward
|   |—— seq2seq
|   |__ topic-classifier
|__ data
	|—— data
	|	|__ utils.py
    |—— convert_to_classification_data_dailydialogue.py
    |—— convert_to_classification_data_empatheticdialogue.py
    |—— process_dailydialogue.py
    |—— process_ELI5.py
    |—— process_empatheticdialogue.py
    |—— process_twitter.py
    |—— process_wikihow.py
    |—— README.md
    |—— topic_stats.py
    |__ topic_stats_combine.py

---

Datasets

Conversational

• DailyDialog

• EmpatheticDialogues

Non-conversational

• Twitter
• Trump Tweets
• Inaugural
• Medical-NLP
• Stanford Sentiment Treebank
• Sentiment140
• ELI5

Experiment Setup

pip install -r requirements.txt

Train

Dialogue Generation

1. Initialization of forward & backward Seq2Seq models

   bash scripts/preprocess_initialization.sh
   bash scripts/initialization.sh
   

Topic Classification

1. Training of topic classifier BERT model

   bash scripts/run_classifier.sh
   

Evaluate