DISK: Domain-constrained Instance Sketch for Math Word Problem Generation

TL;DR

This paper introduces a neural model for generating math word problems from equations, leveraging domain knowledge, instance retrieval, and graph reasoning to produce coherent and contextually accurate problems.

Contribution

It presents a novel approach combining domain-conditioned retrieval and graph-based reasoning to improve math word problem generation from equations.

Findings

Achieves superior performance on automatic evaluation metrics.

Outperforms baseline models in human evaluations.

Enhances understanding of real-world scenarios through graph reasoning.

Abstract

A math word problem (MWP) is a coherent narrative which reflects the underlying logic of math equations. Successful MWP generation can automate the writing of mathematics questions. Previous methods mainly generate MWP text based on inflexible pre-defined templates. In this paper, we propose a neural model for generating MWP text from math equations. Firstly, we incorporate a matching model conditioned on the domain knowledge to retrieve a MWP instance which is most consistent with the ground-truth, where the domain is a latent variable extracted with a domain summarizer. Secondly, by constructing a Quantity Cell Graph (QCG) from the retrieved MWP instance and reasoning over it, we improve the model's comprehension of real-world scenarios and derive a domain-constrained instance sketch to guide the generation. Besides, the QCG also interacts with the equation encoder to enhance the alignment between math tokens (e.g., quantities and variables) and MWP text. Experiments and empirical analysis on educational MWP set show that our model achieves impressive performance in both automatic evaluation metrics and human evaluation metrics.