Hugging Face's logo

Datasets:
CJJones
/
Elementary_Math_Word_Problems_LLM_Training_Short

Tasks:
Question Answering
Text Generation
Modalities:
Text
Formats:
text
Languages:
English
Size:
10K - 100K
Tags:
mathematics
education
word-problems
Libraries:
Datasets
License:
Dataset card Data Studio Files
xet
Community
Elementary_Math_Word_Problems_LLM_Training_Short / README.md
CJJones's picture
CJJones
Update README.md
9bdb22f verified
preview code
|
raw
history blame contribute delete
3.2 kB
metadata
license: cc-by-nc-4.0
language:
  - en
task_categories:
  - question-answering
  - text-generation
tags:
  - mathematics
  - education
  - word-problems
pretty_name: Math Problem Generator Dataset
size_categories:
  - 100K<n<1M

Dataset Card for Math Problem Generator

Dataset Summary

This dataset contains a subset of 100,000 procedurally generated math word problems, covering various mathematical concepts and difficulty levels. The problems were generated using a Java program that creates contextual word problems with detailed solutions and explanations.

🔗 Full dataset available on Gumroad

Supported Tasks and Leaderboards

This dataset can be used for:

  • Math word problem solving
  • Educational AI systems
  • Math question generation
  • Automatic explanation generation

Languages

The dataset is in English.

Dataset Structure

Data Instances

Each instance contains:

  • Problem category/type
  • The word problem text
  • The correct answer
  • A step-by-step explanation

Example: Category: fractions addition Alice has 3/4 of a pizza and finds another 2/5 of the same pizza. How much pizza does Alice have in total? Answer: 23/20 Explanation: To add fractions, find a common denominator (20). Convert 3/4 to 15/20 and 2/5 to 8/20. Add the numerators: 15 + 8 = 23. The answer is 23/20.

text

Data Fields

  • operation: Math category (addition, fractions, algebra, etc.)
  • problem_text: The word problem
  • answer: The correct answer (format varies by problem type)
  • explanation: Step-by-step solution

Data Splits

The dataset comes as a single file with all 100,000 problems.

Dataset Creation

Curation Rationale

This dataset was created to provide a large, diverse set of math word problems for training educational AI systems. The problems cover fundamental math concepts with varying contexts to improve generalization.

Source Data

The problems are procedurally generated using predefined templates and randomization of:

  • Mathematical operations
  • Subjects/objects
  • Character names
  • Real-world contexts

Annotations

Each problem includes:

  • Automatically generated correct answer
  • Step-by-step explanation
  • Problem categorization

Personal and Sensitive Information

The dataset uses common first names and generic objects, containing no real personal information.

Considerations for Using the Data

Social Impact of Dataset

This dataset can help:

  • Develop educational tools for math learning
  • Create AI tutors for students
  • Generate practice problems at scale

Discussion of Biases

Potential biases include:

  • Western-centric names and contexts
  • Limited to elementary/middle school math levels
  • English language only

Other Known Limitations

  • Problems are generated rather than human-written
  • Limited creative variation in problem structures
  • Explanations follow standardized formats

Additional Information

Dataset Curators

C.J. Jones

Licensing Information

cc-by-nc-4.0

Citation Information

If you use this dataset, please cite it as: Part of the C.J. Jones synthetic data collection. With link back to this page.