cnn_dailymail

  • توضیحات :

مجموعه داده های خلاصه سازی غیر ناشناس CNN/DailyMail.

دو ویژگی وجود دارد: - مقاله: متن مقاله خبری که به عنوان سندی برای خلاصه کردن استفاده می شود - نکات برجسته: متن پیوست شده نکات برجسته با و اطراف هر برجسته، که خلاصه هدف است.

  • اسناد اضافی : کاوش در کاغذها با کد

  • صفحه اصلی : https://github.com/abisee/cnn-dailymail

  • کد منبع : tfds.summarization.CnnDailymail

  • نسخه ها :

    • 1.0.0 : API تقسیم جدید ( https://tensorflow.org/datasets/splits )
    • 2.0.0 : جملات هدف را با خط جدید جدا کنید. (با پیش‌بینی مدل جداکننده‌های خط جدید، ارزیابی با استفاده از ROUGE در سطح خلاصه آسان‌تر می‌شود.)

    • 3.0.0 : استفاده از نسخه جعبه ای.

    • 3.1.0 : BuilderConfig حذف شد

    • 3.2.0 : قبل از اضافه شدن دوره جمله، فضای اضافی را حذف کنید. این نباید بر امتیازات ROUGE تأثیر بگذارد زیرا علائم نگارشی حذف شده است.

    • 3.3.0 : اضافه کردن ویژگی ناشر.

    • 3.4.0 (پیش فرض): افزودن ویژگی ID.

  • حجم دانلود : 558.32 MiB

  • حجم مجموعه داده : 1.29 GiB

  • ذخیره خودکار ( اسناد ): خیر

  • تقسیم ها :

شکاف مثال ها
'test' 11,490
'train' 287,113
'validation' 13,368
  • ساختار ویژگی :
FeaturesDict({
    'article': Text(shape=(), dtype=string),
    'highlights': Text(shape=(), dtype=string),
    'id': Text(shape=(), dtype=string),
    'publisher': Text(shape=(), dtype=string),
})
  • مستندات ویژگی :
ویژگی کلاس شکل نوع D شرح
FeaturesDict
مقاله متن رشته
برجسته می کند متن رشته
شناسه متن رشته
ناشر متن رشته
  • نقل قول :
@article{DBLP:journals/corr/SeeLM17,
  author    = {Abigail See and
               Peter J. Liu and
               Christopher D. Manning},
  title     = {Get To The Point: Summarization with Pointer-Generator Networks},
  journal   = {CoRR},
  volume    = {abs/1704.04368},
  year      = {2017},
  url       = {http://arxiv.org/abs/1704.04368},
  archivePrefix = {arXiv},
  eprint    = {1704.04368},
  timestamp = {Mon, 13 Aug 2018 16:46:08 +0200},
  biburl    = {https://dblp.org/rec/bib/journals/corr/SeeLM17},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

@inproceedings{hermann2015teaching,
  title={Teaching machines to read and comprehend},
  author={Hermann, Karl Moritz and Kocisky, Tomas and Grefenstette, Edward and Espeholt, Lasse and Kay, Will and Suleyman, Mustafa and Blunsom, Phil},
  booktitle={Advances in neural information processing systems},
  pages={1693--1701},
  year={2015}
}