clozet/README.md

Cloze Test Dataset

Data Example

{
    "story": 
        "一只猫望着窗外飞翔的鸟儿馋涎欲滴，但自己又捕捉不到。于是它便想了一个法子。它给那些鸟儿们寄去请柬，邀请他们来参加自己的生日宴会。<mask>鸟儿一进来，猫就关上了门。鸟儿们彻底入了虎穴，被猫一只一只抓来吃掉了。", 
    "plot0": 
        "可是没有一只鸟儿愿意来。", 
    "plot1": 
        "有些单纯的鸟儿赴宴来了。", 
    "label": 
        "1"
}

• "story" (str)：input story，<mask> means the removed sentence
• "plot0" (str)：candidate #0
• "plot1" (str)：candidate #1
• "label" (str): 0 means candidate #0 is correct, while 1 means candidate #1 is correct.

Citation

@misc{guan2021lot,
      title={LOT: A Benchmark for Evaluating Chinese Long Text Understanding and Generation}, 
      author={Jian Guan and Zhuoer Feng and Yamei Chen and Ruilin He and Xiaoxi Mao and Changjie Fan and Minlie Huang},
      year={2021},
      eprint={2108.12960},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}

Evaluation

The prediction result should have the same format with test.jsonl

python eval.py prediction_file test.jsonl

We use accuracy as the evaluation metric. The output of the script eval.py is a dictionary as follows:

{"accuracy": _}