mrqa

  • Mô tả :

Nhiệm vụ được chia sẻ MRQA 2019 tập trung vào tính khái quát trong trả lời câu hỏi. Một hệ thống trả lời câu hỏi hiệu quả sẽ làm được nhiều việc hơn là chỉ nội suy từ tập huấn luyện để trả lời các ví dụ kiểm tra được rút ra từ cùng một bản phân phối: nó cũng có thể ngoại suy cho các ví dụ ngoài bản phân phối — một thách thức khó khăn hơn đáng kể.

MRQA điều chỉnh và thống nhất nhiều bộ dữ liệu trả lời câu hỏi riêng biệt (tập hợp con được lựa chọn cẩn thận của bộ dữ liệu hiện có) thành cùng một định dạng (định dạng SQuAD). Trong số đó, sáu bộ dữ liệu được cung cấp để đào tạo và sáu bộ dữ liệu được cung cấp để thử nghiệm. Các phần nhỏ của tập dữ liệu đào tạo được giữ lại dưới dạng dữ liệu trong miền có thể được sử dụng để phát triển. Bộ dữ liệu thử nghiệm chỉ chứa dữ liệu ngoài miền. Điểm chuẩn này được phát hành như một phần của Nhiệm vụ được chia sẻ MRQA 2019.

Bạn có thể tìm thêm thông tin tại: <a href="https://mrqa.github.io/2019/shared.html">https://mrqa.github.io/2019/shared.html</a> .

FeaturesDict({
    'answers': Sequence(string),
    'context': string,
    'context_tokens': Sequence({
        'offsets': int32,
        'tokens': string,
    }),
    'detected_answers': Sequence({
        'char_spans': Sequence({
            'end': int32,
            'start': int32,
        }),
        'text': string,
        'token_spans': Sequence({
            'end': int32,
            'start': int32,
        }),
    }),
    'qid': string,
    'question': string,
    'question_tokens': Sequence({
        'offsets': int32,
        'tokens': string,
    }),
    'subset': string,
})
  • Tài liệu tính năng :
Tính năng Lớp Hình dạng Dtype Sự mô tả
Tính năngDict
câu trả lời Trình tự (Tensor) (Không có,) sợi dây
định nghĩa bài văn tenxơ sợi dây
context_tokens Sự phối hợp
context_tokens/offsets tenxơ int32
context_tokens/tokens tenxơ sợi dây
được phát hiện_answers Sự phối hợp
detect_answers/char_spans Sự phối hợp
detect_answers/char_spans/end tenxơ int32
phát hiện_answers/char_spans/bắt đầu tenxơ int32
phát hiện_answers/văn bản tenxơ sợi dây
detect_answers/token_spans Sự phối hợp
detect_answers/token_spans/end tenxơ int32
detect_answers/token_spans/start tenxơ int32
qid tenxơ sợi dây
câu hỏi tenxơ sợi dây
question_tokens Sự phối hợp
question_tokens/offsets tenxơ int32
question_tokens/tokens tenxơ sợi dây
tập hợp con tenxơ sợi dây

mrqa/squad (cấu hình mặc định)

  • Mô tả cấu hình : Bộ dữ liệu SQuAD (Bộ dữ liệu trả lời câu hỏi Stanford) được sử dụng làm cơ sở cho định dạng tác vụ được chia sẻ. Những người làm việc trong đám đông được xem các đoạn văn từ Wikipedia và được yêu cầu viết câu hỏi với các câu trả lời mang tính khai thác.

  • Kích thước tải xuống : 29.66 MiB

  • Kích thước tập dữ liệu : 271.43 MiB

  • Tự động lưu vào bộ nhớ cache ( tài liệu ): Không

  • Chia tách :

Tách ra ví dụ
'train' 86,588
'validation' 10,507
  • trích dẫn :
@inproceedings{rajpurkar-etal-2016-squad,
    title = "{SQ}u{AD}: 100,000+ Questions for Machine Comprehension of Text",
    author = "Rajpurkar, Pranav  and
      Zhang, Jian  and
      Lopyrev, Konstantin  and
      Liang, Percy",
    booktitle = "Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing",
    month = nov,
    year = "2016",
    address = "Austin, Texas",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D16-1264",
    doi = "10.18653/v1/D16-1264",
    pages = "2383--2392",
}

@inproceedings{fisch-etal-2019-mrqa,
    title = "{MRQA} 2019 Shared Task: Evaluating Generalization in Reading Comprehension",
    author = "Fisch, Adam  and
      Talmor, Alon  and
      Jia, Robin  and
      Seo, Minjoon  and
      Choi, Eunsol  and
      Chen, Danqi",
    booktitle = "Proceedings of the 2nd Workshop on Machine Reading for Question Answering",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-5801",
    doi = "10.18653/v1/D19-5801",
    pages = "1--13",
}

Note that each MRQA dataset has its own citation. Please see the source to see
the correct citation for each contained dataset."

mrqa/news_qa

  • Mô tả cấu hình : Hai nhóm cộng tác viên hỏi và trả lời các câu hỏi dựa trên các bài báo của CNN. “Người hỏi” chỉ xem tiêu đề và tóm tắt của bài viết trong khi “người trả lời” xem toàn bộ bài viết. Các câu hỏi không có câu trả lời hoặc được gắn cờ trong tập dữ liệu là không có thỏa thuận của người chú thích sẽ bị loại bỏ.

  • Kích thước tải xuống : 56.83 MiB

  • Kích thước tập dữ liệu : 654.25 MiB

  • Tự động lưu vào bộ nhớ cache ( tài liệu ): Không

  • Chia tách :

Tách ra ví dụ
'train' 74,160
'validation' 4.212
  • trích dẫn :
@inproceedings{trischler-etal-2017-newsqa,
        title = "{N}ews{QA}: A Machine Comprehension Dataset",
        author = "Trischler, Adam  and
          Wang, Tong  and
          Yuan, Xingdi  and
          Harris, Justin  and
          Sordoni, Alessandro  and
          Bachman, Philip  and
          Suleman, Kaheer",
        booktitle = "Proceedings of the 2nd Workshop on Representation Learning for {NLP}",
        month = aug,
        year = "2017",
        address = "Vancouver, Canada",
        publisher = "Association for Computational Linguistics",
        url = "https://aclanthology.org/W17-2623",
        doi = "10.18653/v1/W17-2623",
        pages = "191--200",
    }
#
@inproceedings{fisch-etal-2019-mrqa,
    title = "{MRQA} 2019 Shared Task: Evaluating Generalization in Reading Comprehension",
    author = "Fisch, Adam  and
      Talmor, Alon  and
      Jia, Robin  and
      Seo, Minjoon  and
      Choi, Eunsol  and
      Chen, Danqi",
    booktitle = "Proceedings of the 2nd Workshop on Machine Reading for Question Answering",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-5801",
    doi = "10.18653/v1/D19-5801",
    pages = "1--13",
}

Note that each MRQA dataset has its own citation. Please see the source to see
the correct citation for each contained dataset."

mrqa/trivia_qa

  • Mô tả cấu hình : Các cặp câu hỏi và câu trả lời được lấy từ các trang web giải đố và đố vui. Phiên bản web của TriviaQA, trong đó ngữ cảnh được truy xuất từ ​​kết quả của truy vấn tìm kiếm Bing, được sử dụng.

  • Kích thước tải xuống : 383.14 MiB

  • Kích thước tập dữ liệu : 772.75 MiB

  • Tự động lưu vào bộ nhớ cache ( tài liệu ): Không

  • Chia tách :

Tách ra ví dụ
'train' 61,688
'validation' 7,785
  • trích dẫn :
@inproceedings{joshi-etal-2017-triviaqa,
    title = "{T}rivia{QA}: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension",
    author = "Joshi, Mandar  and
      Choi, Eunsol  and
      Weld, Daniel  and
      Zettlemoyer, Luke",
    booktitle = "Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2017",
    address = "Vancouver, Canada",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/P17-1147",
    doi = "10.18653/v1/P17-1147",
    pages = "1601--1611",
}

@inproceedings{fisch-etal-2019-mrqa,
    title = "{MRQA} 2019 Shared Task: Evaluating Generalization in Reading Comprehension",
    author = "Fisch, Adam  and
      Talmor, Alon  and
      Jia, Robin  and
      Seo, Minjoon  and
      Choi, Eunsol  and
      Chen, Danqi",
    booktitle = "Proceedings of the 2nd Workshop on Machine Reading for Question Answering",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-5801",
    doi = "10.18653/v1/D19-5801",
    pages = "1--13",
}

Note that each MRQA dataset has its own citation. Please see the source to see
the correct citation for each contained dataset."

mrqa/search_qa

  • Mô tả cấu hình : Các cặp câu hỏi và câu trả lời được lấy từ Jeopardy! Chương trình truyền hình. Các ngữ cảnh bao gồm các đoạn trích được truy xuất từ ​​truy vấn tìm kiếm của Google.

  • Kích thước tải xuống : 699.86 MiB

  • Kích thước tập dữ liệu : 1.38 GiB

  • Tự động lưu vào bộ nhớ cache ( tài liệu ): Không

  • Chia tách :

Tách ra ví dụ
'train' 117,384
'validation' 16.980
  • trích dẫn :
@article{dunn2017searchqa,
    title={Searchqa: A new q\&a dataset augmented with context from a search engine},
    author={Dunn, Matthew and Sagun, Levent and Higgins, Mike and Guney, V Ugur and Cirik, Volkan and Cho, Kyunghyun},
    journal={arXiv preprint arXiv:1704.05179},
    year={2017}
}

@inproceedings{fisch-etal-2019-mrqa,
    title = "{MRQA} 2019 Shared Task: Evaluating Generalization in Reading Comprehension",
    author = "Fisch, Adam  and
      Talmor, Alon  and
      Jia, Robin  and
      Seo, Minjoon  and
      Choi, Eunsol  and
      Chen, Danqi",
    booktitle = "Proceedings of the 2nd Workshop on Machine Reading for Question Answering",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-5801",
    doi = "10.18653/v1/D19-5801",
    pages = "1--13",
}

Note that each MRQA dataset has its own citation. Please see the source to see
the correct citation for each contained dataset."

mrqa/lẩu_qa

  • Mô tả cấu hình : Crowdworkers được hiển thị hai đoạn liên kết thực thể từ Wikipedia và được yêu cầu viết và trả lời các câu hỏi yêu cầu suy luận nhiều bước để giải quyết. Trong cài đặt ban đầu, các đoạn này được trộn lẫn với các đoạn phân tâm bổ sung để làm cho việc suy luận trở nên khó khăn hơn. Ở đây, các đoạn phân tâm không được bao gồm.

  • Kích thước tải xuống : 111.98 MiB

  • Kích thước tập dữ liệu : 272.87 MiB

  • Tự động lưu vào bộ nhớ cache ( tài liệu ): Không

  • Chia tách :

Tách ra ví dụ
'train' 72,928
'validation' 5,901
  • trích dẫn :
@inproceedings{yang-etal-2018-hotpotqa,
    title = "{H}otpot{QA}: A Dataset for Diverse, Explainable Multi-hop Question Answering",
    author = "Yang, Zhilin  and
      Qi, Peng  and
      Zhang, Saizheng  and
      Bengio, Yoshua  and
      Cohen, William  and
      Salakhutdinov, Ruslan  and
      Manning, Christopher D.",
    booktitle = "Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing",
    month = oct # "-" # nov,
    year = "2018",
    address = "Brussels, Belgium",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D18-1259",
    doi = "10.18653/v1/D18-1259",
    pages = "2369--2380",
}

@inproceedings{fisch-etal-2019-mrqa,
    title = "{MRQA} 2019 Shared Task: Evaluating Generalization in Reading Comprehension",
    author = "Fisch, Adam  and
      Talmor, Alon  and
      Jia, Robin  and
      Seo, Minjoon  and
      Choi, Eunsol  and
      Chen, Danqi",
    booktitle = "Proceedings of the 2nd Workshop on Machine Reading for Question Answering",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-5801",
    doi = "10.18653/v1/D19-5801",
    pages = "1--13",
}

Note that each MRQA dataset has its own citation. Please see the source to see
the correct citation for each contained dataset."

mrqa/natural_questions

  • Mô tả cấu hình : Các câu hỏi được thu thập từ các truy vấn tìm kiếm thông tin cho công cụ tìm kiếm Google bởi người dùng thực trong các điều kiện tự nhiên. Các câu trả lời cho các câu hỏi được chú thích trong một trang Wikipedia đã truy xuất bởi cộng đồng. Hai loại chú thích được thu thập: 1) hộp giới hạn HTML chứa đủ thông tin để suy ra hoàn toàn câu trả lời cho câu hỏi (Câu trả lời dài) và 2) khoảng con hoặc các khoảng con trong hộp giới hạn bao gồm câu trả lời thực tế (Câu trả lời ngắn ). Chỉ những ví dụ có câu trả lời ngắn được sử dụng và câu trả lời dài được sử dụng làm ngữ cảnh.

  • Kích thước tải xuống : 121.15 MiB

  • Kích thước tập dữ liệu : 339.03 MiB

  • Tự động lưu vào bộ nhớ cache ( tài liệu ): Không

  • Chia tách :

Tách ra ví dụ
'train' 104,071
'validation' 12,836
  • trích dẫn :
@article{kwiatkowski-etal-2019-natural,
    title = "Natural Questions: A Benchmark for Question Answering Research",
    author = "Kwiatkowski, Tom  and
      Palomaki, Jennimaria  and
      Redfield, Olivia  and
      Collins, Michael  and
      Parikh, Ankur  and
      Alberti, Chris  and
      Epstein, Danielle  and
      Polosukhin, Illia  and
      Devlin, Jacob  and
      Lee, Kenton  and
      Toutanova, Kristina  and
      Jones, Llion  and
      Kelcey, Matthew  and
      Chang, Ming-Wei  and
      Dai, Andrew M.  and
      Uszkoreit, Jakob  and
      Le, Quoc  and
      Petrov, Slav",
    journal = "Transactions of the Association for Computational Linguistics",
    volume = "7",
    year = "2019",
    address = "Cambridge, MA",
    publisher = "MIT Press",
    url = "https://aclanthology.org/Q19-1026",
    doi = "10.1162/tacl_a_00276",
    pages = "452--466",
}

@inproceedings{fisch-etal-2019-mrqa,
    title = "{MRQA} 2019 Shared Task: Evaluating Generalization in Reading Comprehension",
    author = "Fisch, Adam  and
      Talmor, Alon  and
      Jia, Robin  and
      Seo, Minjoon  and
      Choi, Eunsol  and
      Chen, Danqi",
    booktitle = "Proceedings of the 2nd Workshop on Machine Reading for Question Answering",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-5801",
    doi = "10.18653/v1/D19-5801",
    pages = "1--13",
}

Note that each MRQA dataset has its own citation. Please see the source to see
the correct citation for each contained dataset."

mrqa/bio_asq

  • Mô tả cấu hình : BioASQ, một thách thức về lập chỉ mục ngữ nghĩa y sinh quy mô lớn và trả lời câu hỏi, chứa các cặp câu hỏi và câu trả lời được tạo bởi các chuyên gia miền. Sau đó, chúng được liên kết thủ công với nhiều bài báo khoa học (PubMed) liên quan. Bản tóm tắt đầy đủ của từng bài viết được liên kết được tải xuống và sử dụng dưới dạng các ngữ cảnh riêng lẻ (ví dụ: một câu hỏi có thể được liên kết với nhiều bài viết độc lập để tạo ra nhiều cặp ngữ cảnh QA). Các tóm tắt không chứa chính xác câu trả lời sẽ bị loại bỏ.

  • Kích thước tải xuống : 2.54 MiB

  • Kích thước tập dữ liệu : 6.70 MiB

  • Tự động lưu vào bộ đệm ( tài liệu ): Có

  • Chia tách :

Tách ra ví dụ
'test' 1.504
  • trích dẫn :
@article{tsatsaronis2015overview,
    title={An overview of the BIOASQ large-scale biomedical semantic indexing and question answering competition},
    author={Tsatsaronis, George and Balikas, Georgios and Malakasiotis, Prodromos and Partalas, Ioannis and Zschunke, Matthias and Alvers, Michael R and Weissenborn, Dirk and Krithara, Anastasia and Petridis, Sergios and Polychronopoulos, Dimitris and others},
    journal={BMC bioinformatics},
    volume={16},
    number={1},
    pages={1--28},
    year={2015},
    publisher={Springer}
}

@inproceedings{fisch-etal-2019-mrqa,
    title = "{MRQA} 2019 Shared Task: Evaluating Generalization in Reading Comprehension",
    author = "Fisch, Adam  and
      Talmor, Alon  and
      Jia, Robin  and
      Seo, Minjoon  and
      Choi, Eunsol  and
      Chen, Danqi",
    booktitle = "Proceedings of the 2nd Workshop on Machine Reading for Question Answering",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-5801",
    doi = "10.18653/v1/D19-5801",
    pages = "1--13",
}

Note that each MRQA dataset has its own citation. Please see the source to see
the correct citation for each contained dataset."

mrqa/thả

  • Mô tả cấu hình : Các ví dụ về DROP (Lý luận rời rạc về nội dung của Đoạn văn) được thu thập tương tự như SQuAD, trong đó những người làm việc theo nhóm được yêu cầu tạo các cặp câu hỏi-câu trả lời từ các đoạn Wikipedia. Các câu hỏi tập trung vào lý luận định lượng và bộ dữ liệu gốc chứa các câu trả lời dạng số không trích xuất cũng như các câu trả lời dạng văn bản trích xuất. Bộ câu hỏi mang tính chất khai thác được sử dụng.

  • Kích thước tải xuống : 578.25 KiB

  • Kích thước tập dữ liệu : 5.41 MiB

  • Tự động lưu vào bộ đệm ( tài liệu ): Có

  • Chia tách :

Tách ra ví dụ
'test' 1.503
  • trích dẫn :
@inproceedings{dua-etal-2019-drop,
    title = "{DROP}: A Reading Comprehension Benchmark Requiring Discrete Reasoning Over Paragraphs",
    author = "Dua, Dheeru  and
      Wang, Yizhong  and
      Dasigi, Pradeep  and
      Stanovsky, Gabriel  and
      Singh, Sameer  and
      Gardner, Matt",
    booktitle = "Proceedings of the 2019 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)",
    month = jun,
    year = "2019",
    address = "Minneapolis, Minnesota",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/N19-1246",
    doi = "10.18653/v1/N19-1246",
    pages = "2368--2378",
}

@inproceedings{fisch-etal-2019-mrqa,
    title = "{MRQA} 2019 Shared Task: Evaluating Generalization in Reading Comprehension",
    author = "Fisch, Adam  and
      Talmor, Alon  and
      Jia, Robin  and
      Seo, Minjoon  and
      Choi, Eunsol  and
      Chen, Danqi",
    booktitle = "Proceedings of the 2nd Workshop on Machine Reading for Question Answering",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-5801",
    doi = "10.18653/v1/D19-5801",
    pages = "1--13",
}

Note that each MRQA dataset has its own citation. Please see the source to see
the correct citation for each contained dataset."

mrqa/duo_rc

  • Mô tả cấu hình : Phân chia ParaphraseRC của bộ dữ liệu DuoRC được sử dụng. Trong cài đặt này, hai bản tóm tắt cốt truyện khác nhau của cùng một bộ phim được thu thập—một từ Wikipedia và một từ IMDb. Hai nhóm cộng tác viên khác nhau hỏi và trả lời các câu hỏi về cốt truyện phim, trong đó “người hỏi” chỉ hiển thị trên trang Wikipedia và “người trả lời” chỉ hiển thị trên trang IMDb. Các câu hỏi được đánh dấu là không thể trả lời được sẽ bị loại bỏ.

  • Kích thước tải xuống : 1.14 MiB

  • Kích thước tập dữ liệu : 15.04 MiB

  • Tự động lưu vào bộ đệm ( tài liệu ): Có

  • Chia tách :

Tách ra ví dụ
'test' 1.501
  • trích dẫn :
@inproceedings{saha-etal-2018-duorc,
    title = "{D}uo{RC}: Towards Complex Language Understanding with Paraphrased Reading Comprehension",
    author = "Saha, Amrita  and
      Aralikatte, Rahul  and
      Khapra, Mitesh M.  and
      Sankaranarayanan, Karthik",
    booktitle = "Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2018",
    address = "Melbourne, Australia",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/P18-1156",
    doi = "10.18653/v1/P18-1156",
    pages = "1683--1693",
}

@inproceedings{fisch-etal-2019-mrqa,
    title = "{MRQA} 2019 Shared Task: Evaluating Generalization in Reading Comprehension",
    author = "Fisch, Adam  and
      Talmor, Alon  and
      Jia, Robin  and
      Seo, Minjoon  and
      Choi, Eunsol  and
      Chen, Danqi",
    booktitle = "Proceedings of the 2nd Workshop on Machine Reading for Question Answering",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-5801",
    doi = "10.18653/v1/D19-5801",
    pages = "1--13",
}

Note that each MRQA dataset has its own citation. Please see the source to see
the correct citation for each contained dataset."

mrqa/cuộc đua

  • Mô tả cấu hình : Bộ dữ liệu đọc hiểu đọc lại từ các kỳ thi (RACE) được thu thập từ các kỳ thi đọc hiểu tiếng Anh dành cho học sinh trung học cơ sở và trung học phổ thông Trung Quốc. Sự phân chia ở trường trung học (khó khăn hơn) được sử dụng và các câu hỏi kiểu “điền vào chỗ trống” ngầm (không tự nhiên đối với nhiệm vụ này) cũng được lọc ra.

  • Kích thước tải xuống : 1.49 MiB

  • Kích thước tập dữ liệu : 3.53 MiB

  • Tự động lưu vào bộ đệm ( tài liệu ): Có

  • Chia tách :

Tách ra ví dụ
'test' 674
  • trích dẫn :
@inproceedings{lai-etal-2017-race,
    title = "{RACE}: Large-scale {R}e{A}ding Comprehension Dataset From Examinations",
    author = "Lai, Guokun  and
      Xie, Qizhe  and
      Liu, Hanxiao  and
      Yang, Yiming  and
      Hovy, Eduard",
    booktitle = "Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing",
    month = sep,
    year = "2017",
    address = "Copenhagen, Denmark",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D17-1082",
    doi = "10.18653/v1/D17-1082",
    pages = "785--794",
}

@inproceedings{fisch-etal-2019-mrqa,
    title = "{MRQA} 2019 Shared Task: Evaluating Generalization in Reading Comprehension",
    author = "Fisch, Adam  and
      Talmor, Alon  and
      Jia, Robin  and
      Seo, Minjoon  and
      Choi, Eunsol  and
      Chen, Danqi",
    booktitle = "Proceedings of the 2nd Workshop on Machine Reading for Question Answering",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-5801",
    doi = "10.18653/v1/D19-5801",
    pages = "1--13",
}

Note that each MRQA dataset has its own citation. Please see the source to see
the correct citation for each contained dataset."

mrqa/relation_extract

  • Mô tả cấu hình : Với tập dữ liệu lấp đầy vị trí, mối quan hệ giữa các thực thể được chuyển đổi một cách có hệ thống thành các cặp câu trả lời câu hỏi bằng cách sử dụng các mẫu. Ví dụ: mối quan hệ được giáo dục_tại(x, y) giữa hai thực thể x và y xuất hiện trong một câu có thể được diễn đạt là “X được giáo dục tại đâu?” với câu trả lời y. Nhiều mẫu cho từng loại quan hệ được thu thập. Phân chia điểm chuẩn zeroshot của tập dữ liệu (khái quát hóa thành các mối quan hệ không nhìn thấy được) được sử dụng và chỉ các ví dụ tích cực được giữ lại.

  • Kích thước tải xuống : 830.88 KiB

  • Kích thước tập dữ liệu : 3.71 MiB

  • Tự động lưu vào bộ đệm ( tài liệu ): Có

  • Chia tách :

Tách ra ví dụ
'test' 2.948
  • trích dẫn :
@inproceedings{levy-etal-2017-zero,
    title = "Zero-Shot Relation Extraction via Reading Comprehension",
    author = "Levy, Omer  and
      Seo, Minjoon  and
      Choi, Eunsol  and
      Zettlemoyer, Luke",
    booktitle = "Proceedings of the 21st Conference on Computational Natural Language Learning ({C}o{NLL} 2017)",
    month = aug,
    year = "2017",
    address = "Vancouver, Canada",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/K17-1034",
    doi = "10.18653/v1/K17-1034",
    pages = "333--342",
}

@inproceedings{fisch-etal-2019-mrqa,
    title = "{MRQA} 2019 Shared Task: Evaluating Generalization in Reading Comprehension",
    author = "Fisch, Adam  and
      Talmor, Alon  and
      Jia, Robin  and
      Seo, Minjoon  and
      Choi, Eunsol  and
      Chen, Danqi",
    booktitle = "Proceedings of the 2nd Workshop on Machine Reading for Question Answering",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-5801",
    doi = "10.18653/v1/D19-5801",
    pages = "1--13",
}

Note that each MRQA dataset has its own citation. Please see the source to see
the correct citation for each contained dataset."

mrqa/textbook_qa

  • Mô tả cấu hình : Sách giáo khoaQA được thu thập từ các bài học từ sách giáo khoa Khoa học Đời sống, Khoa học Trái đất và Khoa học Vật lý ở trường trung học cơ sở. Các câu hỏi đi kèm với sơ đồ hoặc câu hỏi “Đúng hay Sai” không được bao gồm.

  • Kích thước tải xuống : 1.79 MiB

  • Kích thước tập dữ liệu : 14.04 MiB

  • Tự động lưu vào bộ đệm ( tài liệu ): Có

  • Chia tách :

Tách ra ví dụ
'test' 1.503
  • trích dẫn :
@inproceedings{kembhavi2017you,
    title={Are you smarter than a sixth grader? textbook question answering for multimodal machine comprehension},
    author={Kembhavi, Aniruddha and Seo, Minjoon and Schwenk, Dustin and Choi, Jonghyun and Farhadi, Ali and Hajishirzi, Hannaneh},
    booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern recognition},
    pages={4999--5007},
    year={2017}
}

@inproceedings{fisch-etal-2019-mrqa,
    title = "{MRQA} 2019 Shared Task: Evaluating Generalization in Reading Comprehension",
    author = "Fisch, Adam  and
      Talmor, Alon  and
      Jia, Robin  and
      Seo, Minjoon  and
      Choi, Eunsol  and
      Chen, Danqi",
    booktitle = "Proceedings of the 2nd Workshop on Machine Reading for Question Answering",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-5801",
    doi = "10.18653/v1/D19-5801",
    pages = "1--13",
}

Note that each MRQA dataset has its own citation. Please see the source to see
the correct citation for each contained dataset."