math_qa

  • Descrição :

Um conjunto de dados em grande escala de problemas de palavras matemáticas e um solucionador de problemas de matemática neural interpretável que aprende a mapear problemas para programas de operação.

Dividir Exemplos
'test' 2.985
'train' 29.837
'validation' 4.475
  • Estrutura de recursos :
FeaturesDict({
    'Problem': Text(shape=(), dtype=string),
    'Rationale': Text(shape=(), dtype=string),
    'annotated_formula': Text(shape=(), dtype=string),
    'category': Text(shape=(), dtype=string),
    'correct': Text(shape=(), dtype=string),
    'correct_option': Text(shape=(), dtype=string),
    'linear_formula': Text(shape=(), dtype=string),
    'options': Text(shape=(), dtype=string),
})
  • Documentação do recurso:
Recurso Aula Forma Tipo D Descrição
RecursosDict
Problema Texto corda
Justificativa Texto corda
fórmula_anotada Texto corda
categoria Texto corda
correto Texto corda
opção_correta Texto corda
fórmula_linear Texto corda
opções Texto corda
  • Citação :
@misc{amini2019mathqa,
      title={MathQA: Towards Interpretable Math Word Problem Solving with Operation-Based Formalisms},
      author={Aida Amini and Saadia Gabriel and Peter Lin and Rik Koncel-Kedziorski and Yejin Choi and Hannaneh Hajishirzi},
      year={2019},
      eprint={1905.13319},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}