Ape210K: A Large-Scale and Template-Rich Dataset of Math Word Problems

By Mingyue Shang, Yang Liu, Liang Wang, Jingming Liu. 2020

96 citations recorded by the source corpus.

Automatic math word problem solving has attracted growing attention in recent years. The evaluation datasets used by previous works have serious limitations in terms of scale and diversity. In this paper, we release a new large-scale and template-rich math word problem dataset named Ape210K. It consists of 210K Chinese elementary school-level math problems, which is 9 times the size of the largest public dataset Math23K. Each problem contains both the gold answer and the equations needed to derive the answer. Ape210K is also of greater diversity with 56K templates, which is 25 times more than Math23K. Our analysis shows that solving Ape210K requires not only natural language understanding but also commonsense knowledge. We expect Ape210K to be a benchmark for math word problem solving systems. Experiments indicate that state-of-the-art models on the Math23K dataset perform poorly on Ape210K. We propose a copy-augmented and feature-enriched sequence to sequence (seq2seq) model, which outperforms existing models by 3.2% on the Math23K dataset and serves as a strong baseline of the Ape210K dataset. The gap is still significant between human and our baseline model, calling for further research efforts. We make Ape210K dataset publicly available at this https URL

Subfields: Artificial Intelligence, Information Systems, Natural Language Processing, Evaluation & Benchmarks

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

1. Sergey Levine
2. Quoc V. Le
3. Mohit Bansal
4. Luke S. Zettlemoyer
5. Pieter Abbeel
6. Percy Liang
7. Graham Neubig
8. Yejin Choi

In Information Systems

1. Rui Zhang
2. Qinzhuo Wu
3. Xuanjing Huang
4. Zhongyu Wei
5. Heewoo Jun
6. Wei David Liu
7. John Schulman
8. Qi Zhang

In Natural Language Processing

1. Xuanjing Huang
2. Yue Zhang
3. Cunxiang Wang
4. Rui Zhang
5. Heewoo Jun
6. Mark Chen
7. Wei David Liu
8. Vineet Kosaraju

In Evaluation & Benchmarks

1. Cunxiang Wang
2. Vineet Kosaraju
3. Rui Zhang
4. Christopher Hesse
5. Qinzhuo Wu
6. Xuanjing Huang
7. Jacob Hilton
8. Heewoo Jun

Citations in the corpus

• Training Verifiers to Solve Math Word Problems (2021, background): “Ape210K (Zhao et al., 2020) is the largest publicly available dataset, consisting of 210K Chinese elementary school-level math problems.”
• Math Word Problem Solving with Explicit Numerical Values (2021, methodology): “• We conducted experiments on two large-scale Math23K and Ape210K datasets to verify the effectiveness of our NumS2T model.”
• Exploring Generalization Ability of Pretrained Language Models on Arithmetic and Logical Reasoning (2021, background): “We conduct experiments using BART (Lewis et al., 2020a) since they can generate arbitrary text sequences and have been shown to achieve the state-of-art results on numerous Natural Language Processing (NLP) tasks.”
• CMATH: Can Your Language Model Pass Chinese Elementary School Math Test? (2023, methodology): “, 2017) and APE210K (Zhao et al., 2020) are datasets comprising elementary school level math word problems, which are more similar to our CMATH.”
• A Survey of Deep Learning for Mathematical Reasoning (2022, unknown): “For example, Ape210K (Zhao et al., 2020) consists of 210k elementary math word problems, which is the largest publicly available.”
• Large Language Models for Mathematical Reasoning: Progresses and Challenges (2024, background): “…(Roy and Roth, 2015) 600 E Multi-step reasoning D RAW -1K (Upadhyay and Chang, 2017) 1K E M ATH 23K (Wang et al., 2017) 23K E Chinese A PE 210K (Zhao et al., 2020) 210K E Chinese K6 (Yang et al., 2023) 600 E Chinese; grade 1-6 CM17K (Qin et al., 2021) 17K M H Chinese; grade 6-12 Q u e s ti on -”