Chinese Mining Engineering Knowledge Dataset with 5,287 Q&A Pairs

5,287 Chinese question-answer pairs form a training set for large language models in mining engineering. The dataset, created by a Hefei University of Technology student project, covers six modules including laws, specifications, and safety cases. An enhanced version includes chain-of-thought annotations to improve model reasoning.

Use Cases

• Fine-tune language models on Chinese mining engineering concepts based on the 'concept' module.
• Evaluate model understanding of industrial safety regulations based on the 'law' and 'safety' modules.
• Benchmark model performance on domain-specific reasoning tasks based on the chain-of-thought enhanced dataset.
• Train models to generate technical specifications based on the 'specifications' module.

Strengths

• Contains 5,287 high-quality question-answer pairs for supervised fine-tuning.
• Covers six distinct knowledge modules: laws, specifications, concepts, safety cases, forum experience, and synthesis.
• Offers a separate chain-of-thought enhanced version designed to improve logical reasoning.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the evaluation set is unknown, which may limit suitability assessment.
• Data may reflect the specific focus and source bias inherent to its academic project origin.

Provenance

Source

Hefei University of Technology student innovation project (author: acnul).

Collection Method

Constructed as part of a university-level innovation training program.

Freshness

Last updated 2025-07-24 00:02:53; freshness should be verified.