ServeForEmo
收藏资源简介:
ServeForEmo数据集是由中国科学院自动化研究所多模态人工智能系统国家重点实验室创建,旨在通过模拟真实的情感支持对话来增强大语言模型在情感支持任务中的表现。该数据集包含超过3.7K的多轮对话和62.8K的语句,涵盖了广泛的情感支持场景。数据集的创建过程结合了心理学支持策略,通过角色扮演的方式生成多样化的对话,确保了对话的多样性和有效性。ServeForEmo数据集主要应用于情感支持对话系统,旨在提升系统在处理多样化情感需求时的表现,特别是在心理健康支持和社交互动等领域。
The ServeForEmo dataset was developed by the State Key Laboratory of Multimodal Artificial Intelligence Systems, Institute of Automation, Chinese Academy of Sciences. It aims to enhance the performance of large language models (LLMs) in emotional support tasks by simulating realistic emotional support conversations. This dataset contains over 3.7K multi-turn dialogues and 62.8K utterances, covering a wide range of emotional support scenarios. The dataset was constructed by integrating psychological support strategies, generating diverse dialogues via role-playing to guarantee the diversity and validity of the conversations. The ServeForEmo dataset is primarily applied to emotional support dialogue systems, with the goal of improving the system's performance when handling diverse emotional needs, especially in fields such as mental health support and social interaction.

- 1SweetieChat: A Strategy-Enhanced Role-playing Framework for Diverse Scenarios Handling Emotional Support Agent中国科学院自动化研究所多模态人工智能系统国家重点实验室 · 2024年



