e2e_cleaned

  • คำอธิบาย :

การอัปเดตข้อมูล E2E NLG Challenge พร้อม MR ที่สะอาด ข้อมูล E2E ประกอบด้วยการแสดงความหมายตามบทสนทนา (MR) ในโดเมนร้านอาหารและการอ้างอิงถึง 5 รายการในภาษาธรรมชาติ ซึ่งเป็นสิ่งที่เราจำเป็นต้องคาดเดา

แยก ตัวอย่าง
'test' 4,693
'train' 33,525
'validation' 4,299
  • โครงสร้างคุณลักษณะ :
FeaturesDict({
    'input_text': FeaturesDict({
        'table': Sequence({
            'column_header': string,
            'content': string,
            'row_number': int16,
        }),
    }),
    'target_text': string,
})
  • เอกสารคุณสมบัติ :
ลักษณะเฉพาะ ระดับ รูปร่าง Dประเภท คำอธิบาย
คุณสมบัติDict
อินพุต_ข้อความ คุณสมบัติDict
input_text/ตาราง ลำดับ
input_text/table/column_header เทนเซอร์ สตริง
input_text/table/content เทนเซอร์ สตริง
input_text/table/row_number เทนเซอร์ int16
target_text เทนเซอร์ สตริง
  • การอ้างอิง :
@inproceedings{dusek-etal-2019-semantic,
    title = "Semantic Noise Matters for Neural Natural Language Generation",
    author = "Du{\v{s} }ek, Ond{\v{r} }ej  and
      Howcroft, David M.  and
      Rieser, Verena",
    booktitle = "Proceedings of the 12th International Conference on Natural Language Generation",
    month = oct # "{--}" # nov,
    year = "2019",
    address = "Tokyo, Japan",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/W19-8652",
    doi = "10.18653/v1/W19-8652",
    pages = "421--426",
    abstract = "Neural natural language generation (NNLG) systems are known for their pathological outputs, i.e. generating text which is unrelated to the input specification. In this paper, we show the impact of semantic noise on state-of-the-art NNLG models which implement different semantic control mechanisms. We find that cleaned data can improve semantic correctness by up to 97{\%}, while maintaining fluency. We also find that the most common error is omitting information, rather than hallucination.",
}