Techniques to Improve Neural Math Word Problem Solvers

TL;DR

This paper introduces a new neural network architecture for solving math word problems that better utilizes question text and enforces mathematical laws, leading to improved performance over existing methods.

Contribution

It proposes a novel encoder-decoder model that incorporates question text and uses Deep Sets to ensure permutation invariance, enhancing reasoning in neural MWP solvers.

Findings

Outperforms state-of-the-art neural MWP solvers on four benchmarks.

Effectively leverages question text to improve understanding.

Enforces mathematical laws to produce more accurate expressions.

Abstract

Developing automatic Math Word Problem (MWP) solvers is a challenging task that demands the ability of understanding and mathematical reasoning over the natural language. Recent neural-based approaches mainly encode the problem text using a language model and decode a mathematical expression over quantities and operators iteratively. Note the problem text of a MWP consists of a context part and a question part, a recent work finds these neural solvers may only perform shallow pattern matching between the context text and the golden expression, where question text is not well used. Meanwhile, existing decoding processes fail to enforce the mathematical laws into the design, where the representations for mathematical equivalent expressions are different. To address these two issues, we propose a new encoder-decoder architecture that fully leverages the question text and preserves step-wise commutative law. Besides generating quantity embeddings, our encoder further encodes the question text and uses it to guide the decoding process. At each step, our decoder uses Deep Sets to compute expression representations so that these embeddings are invariant under any permutation of quantities. Experiments on four established benchmarks demonstrate that our framework outperforms state-of-the-art neural MWP solvers, showing the effectiveness of our techniques. We also conduct a detailed analysis of the results to show the limitations of our approach and further discuss the potential future work. Code is available at https://github.com/sophistz/Question-Aware-Deductive-MWP.