Adversarial Examples for Evaluating Math Word Problem Solvers

Standard accuracy metrics have shown that Math Word Problem (MWP) solvers have achieved high performance on benchmark datasets. However, the extent to which existing MWP solvers truly understand language and its relation with numbers is still unclear. In this paper, we generate adversarial attacks t...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2021-09
Main Authors: Kumar, Vivek, Maheshwary, Rishabh, Pudi, Vikram
Format: Article
Language:English
Subjects:
Benchmarks
Datasets
Solvers
Words (language)
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Standard accuracy metrics have shown that Math Word Problem (MWP) solvers have achieved high performance on benchmark datasets. However, the extent to which existing MWP solvers truly understand language and its relation with numbers is still unclear. In this paper, we generate adversarial attacks to evaluate the robustness of state-of-the-art MWP solvers. We propose two methods Question Reordering and Sentence Paraphrasing to generate adversarial attacks. We conduct experiments across three neural MWP solvers over two benchmark datasets. On average, our attack method is able to reduce the accuracy of MWP solvers by over 40 percentage points on these datasets. Our results demonstrate that existing MWP solvers are sensitive to linguistic variations in the problem text. We verify the validity and quality of generated adversarial examples through human evaluation.
ISSN:2331-8422