ibokajordan/turkish_metaverse_eco_dialogues

🧾 Dataset Card for turkishmetaverseeco_dialogues

Dataset Summary

The turkish_metaverse_eco_dialogues dataset is a Turkish-language corpus developed for Retrieval-Augmented Generation (RAG) and instruction fine-tuning tasks. It contains natural, context-rich dialogues centered around three thematic areas — daily life, metaverse, and eco-friendly communication — designed to train and evaluate large language models (LLMs) for knowledge-grounded, context-aware, and goal-oriented text generation.

Supported Tasks and Benchmarks

• —Text Generation – producing fluent, coherent, and contextually relevant Turkish responses.
• —Question Answering (QA) – retrieving and reasoning over contextual information.
• —Instruction Fine-Tuning – adapting pre-trained LLMs to Turkish domain-specific dialogue styles.
• —RAG Evaluation – serving as a benchmark for context retrieval and generation quality in low-resource languages.

Languages

• —Primary language: Turkish (tr)
• —Script: Latin
• —Register: Conversational (informal, natural, and colloquial Turkish)

Dataset Structure

Each entry in the dataset consists of a question, context, and answer field. The data is provided in both .jsonl and .csv formats to support diverse training pipelines.

Example:

{
  "question": "What is metaverse?",
  "context": "The metaverse is like a virtual world within the internet.",
  "answer": "It is a type of online universe."
}

Data Splits

Domains and Semantic Coverage

Licensing Information

• —License: Creative Commons Attribution 4.0 International (CC-BY-4.0) You are free to share and adapt the dataset, provided attribution is given to the author.

Citation

If you use this dataset, please cite:

Ozkal, i. (2025). turkish_metaverse_eco_dialogues [Dataset]. Hugging Face. https://huggingface.co/datasets/ibokajordan/turkishmetaverseeco_dialogues

APA Format: Ozkal, İ. (2025). turkish_metaverse_eco_dialogues: Turkish Dialogues for Retrieval-Augmented Generation and Fine-Tuning [Dataset]. Hugging Face. https://huggingface.co/datasets/ibokajordan/turkishmetaverseeco_dialogues

Ethical Considerations

• —The dataset does not contain personal or sensitive information.
• —All dialogues were manually curated or generated for educational and research use.
• —Toxic, biased, or discriminatory language has been filtered.
• —Distributed responsibly to support transparency and inclusivity in AI research.

Dataset Curators

Developed by Ozkal (2025) for academic research on Turkish-language human–AI communication and contextual LLM alignment.