cs_restaurants

  • Описание :

Чешский набор данных для преобразования данных в текст в домене ресторана. Входные смысловые представления содержат тип диалогового действия (информировать, подтвердить и т. д.), слоты (еда, площадь и т. д.) и их значения. Он возник как перевод англоязычного набора данных о ресторанах Сан-Франциско, сделанный Wen et al. (2015).

Расколоть Примеры
'test' 842
'train' 3569
'validation' 781
  • Структура функции :
FeaturesDict({
    'delex_input_text': FeaturesDict({
        'table': Sequence({
            'column_header': string,
            'content': string,
            'row_number': int16,
        }),
    }),
    'delex_target_text': string,
    'input_text': FeaturesDict({
        'table': Sequence({
            'column_header': string,
            'content': string,
            'row_number': int16,
        }),
    }),
    'target_text': string,
})
  • Документация по функциям :
Особенность Учебный класс Форма Dтип Описание
ОсобенностиDict
delex_input_text ОсобенностиDict
delex_input_text/таблица Последовательность
delex_input_text/таблица/column_header Тензор нить
delex_input_text/таблица/контент Тензор нить
delex_input_text/таблица/номер_строки Тензор int16
delex_target_text Тензор нить
input_text ОсобенностиDict
input_text/таблица Последовательность
input_text/таблица/column_header Тензор нить
input_text/таблица/содержание Тензор нить
input_text/таблица/номер_строки Тензор int16
целевой_текст Тензор нить
  • Цитата :
@inproceedings{dusek_neural_2019,
        author = {Dušek, Ondřej and Jurčíček, Filip},
        title = {Neural {Generation} for {Czech}: {Data} and {Baselines} },
        shorttitle = {Neural {Generation} for {Czech} },
        url = {https://www.aclweb.org/anthology/W19-8670/},
        urldate = {2019-10-18},
        booktitle = {Proceedings of the 12th {International} {Conference} on {Natural} {Language} {Generation} ({INLG} 2019)},
        month = oct,
        address = {Tokyo, Japan},
        year = {2019},
        pages = {563--574},
        abstract = {We present the first dataset targeted at end-to-end NLG in Czech in the restaurant domain, along with several strong baseline models using the sequence-to-sequence approach. While non-English NLG is under-explored in general, Czech, as a morphologically rich language, makes the task even harder: Since Czech requires inflecting named entities, delexicalization or copy mechanisms do not work out-of-the-box and lexicalizing the generated outputs is non-trivial. In our experiments, we present two different approaches to this this problem: (1) using a neural language model to select the correct inflected form while lexicalizing, (2) a two-step generation setup: our sequence-to-sequence model generates an interleaved sequence of lemmas and morphological tags, which are then inflected by a morphological generator.},
}