Deep Neural Solver for Math Word Problems

By Yan Wang, Xiaojiang Liu, Shuming Shi. 2017

397 citations recorded by the source corpus.

This paper presents a deep neural solver to automatically solve math word problems. In contrast to previous statistical learning approaches, we directly translate math word problems to equation templates using a recurrent neural network (RNN) model, without sophisticated feature engineering. We further design a hybrid model that combines the RNN model and a similarity-based retrieval model to achieve additional performance improvement. Experiments conducted on a large dataset show that the RNN model and the hybrid model significantly outperform state-of-the-art statistical learning methods for math word problem solving.

Subfields: Artificial Intelligence, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Natural Language Processing

Citations in the corpus

NumNet: Machine Reading Comprehension with Numerical Reasoning (2019, neutral): “Wang et al. (2017b) and Ling et al. (2017) proposed sequence to sequence solvers for the AWP problems, which are capable of generating unseen expressions and do not rely on sophisticated manual features.”
Training Verifiers to Solve Math Word Problems (2021, background): “Previous work has attempted to solve classic math word problem benchmarks with recurrent seq2seq models (Sutskever et al., 2014) and closely related variants (Wang et al., 2017; Huang et al., 2018).”
MWPToolkit: An Open-Source Framework for Deep Learning-Based Math Word Problem Solvers (2021, methodology): “, 2016) Draw1K en Multiple equations generation 1,000 745 - (Upadhyay and Chang, 2017) Math23K zh Single equation generation 23,162 (Wang et al., 2017) HMWP zh Multiple equations generation 5,491 1,789 (Qin et al.”
Are NLP Models really able to Solve Simple Math Word Problems? (2021, methodology): “rated the MAWPS dataset from previous datasets which along with Math23k (Wang et al., 2017) has been used as benchmark in recent works.”
MathQA: Towards Interpretable Math Word Problem Solving with Operation-Based Formalisms (2019, background): “Neural Word Problem Solvers Following the increasing availability of large-scale datasets like AQuA, several recent works have explored deep neural approaches to math word problem solving (Wang et al., 2017).”
Learning to Reason Deductively: Math Word Problem Solving as Complex Relation Extraction (2022, methodology): “Chinese BERT and Chinese Roberta (Cui et al., 2019) are used for Math23k.”
Generate & Rank: A Multi-task Framework for Math Word Problems (2021, positive): “We conduct the experiments on two commonly-used datasets: Math23K (Wang et al., 2017) and MAWPS (Koncel-Kedziorski et al., 2016).”
A Diverse Corpus for Evaluating and Developing English Math Word Problem Solvers (2020, background): “(addition, subtraction, multiplication and division) with either single-step or multi-step operations; (3) the Algebraic Word Problem corpora (Kushman et al., 2014; Koncel-Kedziorski et al., 2015; Roy and Roth, 2017; Upadhyay & Chang, 2015; Wang et al., 2017), which focus on algebraic MWPs; and (4) ”
Semantically-Aligned Equation Generation for Solving and Reasoning Math Word Problems (2018, methodology): “The experiments are benchmarked on the dataset Math23k (Wang et al., 2017), which contains 23,162 math problems with annotated equations.”
Analysing Mathematical Reasoning Abilities of Neural Models (2019, background): “These questions tend to be crowd sourced or obtained from exercise books, and existing datasets include Allen Institute for AI (2014); Kushman et al. (2014); Huang et al. (2016); Upadhyay & Chang (2016); Wang et al. (2017); Ling et al. (2017).”
A Goal-Driven Tree-Structured Neural Model for Math Word Problems (2019, methodology): “Researchers have proposed several approaches [Ling et al. , 2017; Wang et al. , 2017; Huang et al. , 2018; Wang et al. , 2018a] to solving MWPs based on the Seq2Seq model [Sutskever et al. , 2014].”
MathDQN: Solving Arithmetic Word Problems via Deep Reinforcement Learning (2018, background): “An initial study in this area can be found in (Wang et al. 2017; Huang et al. 2017).”
QuaRel: A Dataset and Models for Answering Questions about Qualitative Relationships (2018, background): “Answering story problems has received attention in the domain of arithmetic, where simple algebra story questions (e.g., “Sue had 5 cookies, then gave 2 to Joe...”) are mapped to a system of equations, e.g., (Ling et al. 2017; Kushman et al. 2014; Wang, Liu, and Shi 2017; Shi et al. 2015).”
Modeling Intra-Relation in Math Word Problems with Different Functional Multi-Head Attentions (2019, methodology): “In recent years, a growing number of deep learning models for MWPs (Wang et al., 2017; Ling et al., 2017; Wang et al., 2018b,a; Huang et al., 2018a,b; Wang et al., 2019) have drawn inspiration from advances in machine translation.”
Template-Based Math Word Problem Solvers with Recursive Neural Networks (2019, methodology): “Among the seq2seq generation models, DNS performs much better than simple LSTM or CNN.”
Graph-to-Tree Learning for Solving Math Word Problems (2020, methodology): “Wang et al. (2017) proposed a large-scale MWP dataset and applied a vanilla sequence to sequence (seq2seq) model to translate the language text to a solution expression.”
A Meaning-Based Statistical English Math Word Problem Solver (2018, methodology): “, 2016), in which all related decisions are done via a statistical classifier; (3) DNNbased approaches (Ling et al., 2017; Wang et al., 2017), which map the given text into the corresponding math operation/equation via a DNN; and (4) Mixed approaches, which identify entities and quantities with rule”
GeoQA: A Geometric Question Answering Benchmark Towards Multimodal Numerical Reasoning (2021, background): “…problems, including traditional machine learning approaches (Kushman et al., 2014; Zhou et al., 2015; Huang et al., 2016) and network-based models (Wang et al., 2017, 2018; Xie and Sun, 2019), while solving geometric problems has been rarely investigated (Seo et al., 2014, 2015a; Sachan et al.,…”
A Weakly Supervised Model for Solving Math word Problems (2021, methodology): “Hybrid model w/ SNI (Wang et al., 2017) is a combination of the retrieval model and the RNN-based Seq2Seq model with signiﬁcant number iden-tiﬁcation (SNI).”