Turkish Multi-Domain Customer Service Dialogues in ChatML Format

Turkish conversations between customers and agents across sectors like finance, e-commerce, and technical support. This is a version of the original 'Turkish Call Center Conversations' dataset from Kaggle, converted into ChatML format for fine-tuning large language models. The dataset was uploaded by emreseyhan and was last updated on Hugging Face in May 2026.

Use Cases

• Fine-tuning large language models for Turkish customer service tasks based on the described ChatML format.
• Training intent classification models based on the multi-domain nature of the conversations (finance, e-commerce, technical support).
• Developing dialogue systems for Turkish-language technical support based on the described conversation types.
• Analyzing Turkish customer-agent interaction patterns across different business sectors mentioned in the description.

Strengths

• Conversations cover multiple business domains as described, including finance and e-commerce.
• Data is formatted specifically for LLM fine-tuning in ChatML format.
• Dataset is derived from an original Kaggle dataset, suggesting a prior level of curation.

Limitations

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

Provenance

Source

Original dataset 'Turkish Call Center Conversations' by Anıl Sevinç on Kaggle.

Collection Method

Converted from the original Kaggle dataset into ChatML format.

Freshness

Last updated 2026-05-18 16:33:40; freshness should be verified.

Geography

Turkey (inferred from language and original source).