SocratDataset-EN: English Elementary Science Tutoring Dialogues

An English translation of a Chinese corpus for training Socratic teaching models. The dataset was created by ulises-c and last updated on May 4, 2026. It enables English-language research and fine-tuning without requiring access to the original Chinese data.

Use Cases

• Fine-tuning large language models for Socratic teaching based on the described tutoring dialogue structure
• Researching pedagogical dialogue patterns in elementary science education based on the described tutoring framework
• Developing English-language educational AI assistants based on the translated dialogue corpus

Strengths

• Complete English translation of the original Chinese training corpus, enabling broader access.
• Specifically designed for training Socratic teaching models, following the SocRule framework.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

Translation of ulises-c/SocratDataset, the training corpus for SocratTeachLLM (KELE, EMNLP 2025 Findings).

Collection Method

English translation of Chinese elementary-school science tutoring dialogues.

Freshness

Last updated 2026-05-04 07:57:40; freshness should be verified.