热点信息问答语料

覆盖重大以及网民关注的热点问题。结合原始语料、重点领域语料，并对比境外媒体数据，通过人机结合的方式形成多样化的问答数据。关于同一内容，可以形成不同的任务指令提示，提出不同的问题表述，归纳总结形成不同视角、不同风格的应答内容。针对重大问题问答，为支撑生成式AI的安全开展，为大模型提供正确价值导向。This dataset covers major issues and hot topics of public concern among netizens. By combining original corpora and domain-specific key corpora, and comparing with overseas media data, diversified question-answering data is generated through a human-machine collaborative approach. For the same content, distinct task instruction prompts can be designed, varied question formulations can be proposed, and response content with different perspectives and styles can be summarized and created. For question-answering tasks on major issues, this dataset supports the safe deployment of generative AI and provides correct value orientation for large language models (LLMs).

搜集汇总

背景与挑战

背景概述

该数据集为热点信息问答语料，涵盖重大社会热点和网民关注问题，通过人机协作生成多样化问答内容，旨在为生成式AI提供安全可靠的价值导向训练素材。