MathQA: Towards Interpretable Math Word Problem Solving with Operation-Based Formalisms
Aida Amini University of Washington Saadia Gabriel University of Washington Shanchuan Lin University of Washington Rik Koncel-Kedziorski University of Washington Yejin Choi University of Washington Allen Institute for AI Hannaneh Hajishirzi University of Washington Allen Institute for AI
Abstract
We introduce a large-scale dataset of math word problems and an interpretable neural math problem solver that learns to map problems to operation programs. Due to annotation challenges, current datasets in this domain have been either relatively small in scale or did not offer precise operational annotations over diverse problem types. We introduce a new representation language to model precise operation programs corresponding to each math problem that aim to improve both the performance and the interpretability of the learned models. Using this representation language, our new dataset, MathQA, significantly enhances the AQuA dataset with fully-specified operational programs. We additionally introduce a neural sequence-to-program model enhanced with automatic problem categorization. Our experiments show improvements over competitive baselines in our MathQA as well as the AQuA datasets. The results are still significantly lower than human performance indicating that the dataset poses new challenges for future research. Our dataset is available at: https://math-qa.github.io/math-QA/.
中文速览
数学应用题的自动求解长期面临两大瓶颈：现有数据集要么规模太小，要么缺乏精确的逐步运算标注，导致模型既难学好、又难以解释推理过程。研究者设计了一套包含58种运算符的操作程序表示语言（operation program representation language），用它对大规模英文数学选择题数据集AQuA进行重新标注，构建出拥有3.7万道题的MathQA数据集，每道题都配有可逐步执行的形式化运算程序。在此基础上，他们提出了一个结合自动题目分类的神经序列到程序（sequence-to-program）模型，实验结果在MathQA和AQuA两个测试集上均超越了此前最优方法。不过模型表现仍与人类水平存在明显差距，说明该数据集为数学推理研究提供了真正有挑战性的新基准，对推动可解释的自动数学解题具有重要价值。
原文 arXiv:1905.13319;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1905.13319v1