Benchmark Dataset for OR Problems
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/bwz96sco/or_llm_agent
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含83个以自然语言描述的实际操作研究（OR）问题，每个问题都附有以LaTeX或Markdown格式编排的必要表格数据。该数据集作为评估人工智能模型解决方案准确性的基准，并已通过手动整理和验证。规模覆盖了83个实际OR问题，其任务旨在评估自动化OR求解代理在自然语言描述的OR问题上的性能表现。
This dataset contains 83 real-world Operations Research (OR) problems described in natural language, each accompanied by necessary tabular data formatted in LaTeX or Markdown. Manually curated and validated, this dataset serves as a benchmark for evaluating the accuracy of solutions generated by artificial intelligence models, and is designed to assess the performance of automated OR-solving AI Agents on natural language-described OR problems.
