llm-semantic-router/feedback-detector-dataset
收藏Hugging Face2026-01-21 更新2026-02-07 收录
下载链接：
https://hf-mirror.com/datasets/llm-semantic-router/feedback-detector-dataset
下载链接
链接失效反馈官方服务：
资源简介：
这是一个大规模多语言用户反馈分类数据集，包含51,694个样本，分为4个类别：满意（SAT）、需要澄清（NEED_CLARIFICATION）、错误答案（WRONG_ANSWER）和想要不同（WANT_DIFFERENT）。数据集来源于多个公开的对话和投诉数据集，包括英语、日语和土耳其语。标注过程使用了OpenAI GPT-OSS-120B模型在AMD MI300X GPU上完成，具有确定性输出、结构化JSON输出、重试逻辑和并行处理等特点。数据集适用于微调反馈检测模型、用户满意度分类、客户服务自动化和对话系统评估等用途。
A large-scale multilingual dataset for 4-class user feedback classification, containing 51,694 examples labeled into SAT (satisfied), NEED_CLARIFICATION, WRONG_ANSWER, and WANT_DIFFERENT. The dataset combines multiple public dialogue and complaint datasets in English, Japanese, and Turkish. Labels were generated using OpenAI GPT-OSS-120B on AMD MI300X GPU with deterministic output, structured JSON, retry logic, and parallel processing. Intended for fine-tuning feedback detection models, user satisfaction classification, customer service automation, and dialogue system evaluation.
提供机构：
llm-semantic-router
