sunorme/chinese-adorable-high-emotional-intelligence-chat
收藏资源简介：
# 🩷 Chinese Adorable High Emotional Intelligence Chat Dataset ### 💬 中文高情商可爱聊天数据集 ## 简要参数 --- license: cc-by-4.0 task_categories: - table-question-answering language: - zh tags: - chat - emotional size_categories: - n<1K --- ## 🧩 简介 (Overview) **Chinese Adorable High Emotional Intelligence Chat Dataset** 是一个中文对话数据集，专注于**高情商、轻松幽默、温柔治愈风格**的自然对话。 对话以“**user**”和“**girl**”为角色构成，模拟出一种**温柔、聪慧且带点俏皮的女性语气**，用于训练能自然、情绪感知良好的中文对话模型。 本数据集尤其适合： * 微调情绪对话模型（Emotional Chatbot） * 训练高情商人格角色（Roleplay / Companion AI） * 情绪识别与反应生成任务（Emotion-aware Response Generation） * 人格化对话研究（Persona-based Dialogue） --- ## 📦 数据示例 (Data Example) ```json { "user": "你今天状态不太好。", "girl": "那我今天就当你的充电宝，抱一下回血百分之十～" } ``` 或另一组更完整的上下文： ```json [ { "user": "你不是说要早起跑步吗？", "girl": "是啊，但床太粘人了，我一挣扎它就更抱我紧了～" }, { "user": "真拿你没办法。", "girl": "那你就别拿了，直接抱吧～" } ] ``` --- ## 🧠 数据特征 (Characteristics) | 特征 | 描述 | | -------- | ----------------------------------- | | **语言** | 中文 | | **数据量** | 约 170 组高情商对话 | | **格式** | JSON 数组，每条包含 `"user"` 与 `"girl"` 字段 | | **对话风格** | 温柔、调皮、聪明、善于共情 | | **情绪维度** | 轻松 / 暧昧 / 安慰 / 打趣 / 撩动 / 理解 | | **话语长度** | 平均 15–30 个字，语言自然且含情绪转折 | --- ## 💡 设计理念 (Design Philosophy) 本数据集的核心目标是捕捉“**高情商语用模式**”的语言特征，包括： * **共情性回应**：理解情绪并以轻松语气化解； * **语言柔化**：使用“拟人、调侃、撒娇”等表达降低冲突； * **人际亲密语气**：模糊距离感的表达，如“抱”“夸”“小笨蛋”等； * **隐含心理反馈**：以轻喜剧语态传达情绪调节能力。 这种语言风格常见于高情绪智商的社交互动、伴侣式AI或轻治愈向角色中。 --- ## 🧰 数据格式 (Schema) 每条数据均为独立对话对象，格式如下： | 字段 | 类型 | 描述 | | ------ | -------- | ------------------- | | `user` | `string` | 用户发言（通常为中性语气或情境引导） | | `girl` | `string` | 高情商女性角色回复，富有情感与语言技巧 | 示例： ```json { "user": "你生气了吗？", "girl": "没有呀～只是心情暂时进了小黑屋，等你一句哄就能放风～" } ``` --- ## 🪄 使用场景 (Use Cases) | 场景 | 示例 | | ---------------------- | ---------------- | | 🎭 **情感陪伴AI / 虚拟女友模型** | 微调以获得自然的暧昧与安慰语气 | | 🧘 **情绪引导与安抚模型** | 生成具同理心的回应 | | 💬 **对话生成研究** | 用于风格迁移 / 情感标签学习 | | 💞 **个性化语料训练** | 构建“温柔型人格”Chatbot | --- ## ⚖️ 许可协议 (License) 本数据集建议使用： **🪪 License:**[CC BY 4.0 (Attribution)](https://creativecommons.org/licenses/by/4.0/) > 允许自由研究、修改与分发。 > 请在使用时注明出处： > > ``` > Dataset: Chinese Adorable High Emotional Intelligence Chat (2025) > Author: memorialsummer > ``` --- ## 🏗️ 未来扩展 (Future Work) * 增加更多语气类型（治愈 / 犀利 / 理性 / 哄劝） * 加入多轮上下文对话（Context-aware） * 添加情感标签（如 joy / comfort / tease） * 引入语音语料（Tone-aware fine-tuning） --- ## 🌸 引用 (Citation) 如果你在研究或项目中使用本数据集，请引用如下格式： ```bibtex @dataset{memorialsummer_2025adorablechat, title = {Chinese Adorable High Emotional Intelligence Chat Dataset}, author = {Memorial_Summer}, year = {2025}, note = {https://huggingface.co/datasets/MemorialSummer/chinese-adorable-high-emotional-intelligence-chat} } ```
# 🩷 Chinese Adorable High Emotional Intelligence Chat Dataset ### 💬 Chinese High-EQ Cute Chat Dataset ## Key Parameters --- license: cc-by-4.0 task_categories: - table-question-answering language: - zh tags: - chat - emotional size_categories: - n<1K --- ## 🧩 Overview **Chinese Adorable High Emotional Intelligence Chat Dataset** is a Chinese conversational dataset focused on natural dialogues characterized by high emotional intelligence, lighthearted humor, and gentle healing tones. The dialogues are structured with two roles: "user" and "girl", which simulate a gentle, witty, and slightly playful female tone, designed to train Chinese conversational models capable of natural, emotionally-aware interactions. This dataset is particularly suitable for: * Fine-tuning emotional chatbots * Training high-EQ persona-based roles (Roleplay / Companion AI) * Emotion-aware response generation tasks * Persona-based dialogue research --- ## 📦 Data Examples Single-turn dialogue example: json { "user": "You don't seem to be yourself today.", "girl": "Then I'll be your power bank today. A hug will recharge you by 10%~" } Or a complete multi-turn context example: json [ { "user": "Didn't you say you were going to get up early for a run?", "girl": "Yeah, but the bed is so clingy, it hugged me tighter every time I tried to get up~" }, { "user": "I really have no way with you.", "girl": "Then don't deal with me, just hug me instead~" } ] --- ## 🧩 Dataset Characteristics | Characteristic | Description | | -------------------- | -------------------------------------------- | | **Language** | Chinese | | **Dataset Size** | Approximately 170 high-EQ dialogue pairs | | **Format** | JSON array, each entry contains "user" and "girl" fields | | **Dialogue Style** | Gentle, playful, witty, and empathetic | | **Emotion Dimensions** | Lighthearted / flirtatious / comforting / teasing / flirty / understanding | | **Utterance Length** | Average 15-30 words, natural language with emotional twists | --- ## 💡 Design Philosophy The core goal of this dataset is to capture the linguistic features of "high-EQ pragmatic patterns", including: * **Empathetic Responses**: Understanding emotions and diffusing tension with a lighthearted tone; * **Linguistic Softening**: Using expressions like personification, teasing, and coquetry to reduce conflicts; * **Intimate Interpersonal Tone**: Expressions that blur social distance, such as "hug", "praise", "silly", etc.; * **Implicit Psychological Feedback**: Conveying emotional regulation abilities through light comedic delivery. This linguistic style is commonly seen in high-EQ social interactions, companion AI, or light healing-oriented roles. --- ## 🧰 Data Schema Each entry is an independent dialogue object with the following format: | Field | Type | Description | | ------ | ---------- | -------------------------------------------- | | `user` | `string` | User's utterance (usually neutral tone or situational prompt) | | `girl` | `string` | High-EQ female character's response, rich in emotion and linguistic skill | Example: json { "user": "Are you mad?", "girl": "No~ I just temporarily locked my mood in a small dark room, it will let out once you coax me~" } --- ## 🪄 Use Cases | Use Case | Example | | ----------------------------------------- | ----------------------------------------------------------------------- | | 🎭 **Emotional Companion AI / Virtual Girlfriend Model** | Fine-tune to acquire natural flirtatious and comforting tones | | 🧘 **Emotional Guidance & Comfort Model** | Generate empathetic responses | | 💬 **Dialogue Generation Research** | Used for style transfer / emotion label learning | | 💞 **Personalized Corpus Training** | Build "gentle persona" chatbots | --- ## ⚖️ License This dataset is recommended to be used under: **🪪 License:** [CC BY 4.0 (Attribution)](https://creativecommons.org/licenses/by/4.0/) > Free research, modification, and distribution are permitted. > Please cite the source when using: > > > Dataset: Chinese Adorable High Emotional Intelligence Chat (2025) > Author: memorialsummer > --- ## 🏗️ Future Work * Add more tone types (healing / sharp / rational / comforting) * Add multi-turn context-aware dialogues * Add emotion tags (e.g., joy / comfort / tease) * Introduce speech corpora for tone-aware fine-tuning --- ## 🌸 Citation If you use this dataset in your research or project, please cite the following format: bibtex @dataset{memorialsummer_2025adorablechat, title = {Chinese Adorable High Emotional Intelligence Chat Dataset}, author = {Memorial_Summer}, year = {2025}, note = {https://huggingface.co/datasets/MemorialSummer/chinese-adorable-high-emotional-intelligence-chat} }
