教科书级问答语料

高质量的提问和回答对应文本语料。分文科、理工科、医学、综合四个类别，包含金融、医疗、法律、电力、工业、机械、维修、生活、百科等20余个专业领域的知识问答语料。包括知识性一问一答数据和场景性连续多轮问答数据等形式。可直接用于人工智能自然语言处理的研发和应用、大语言模型研发的训练、AIGC应用的训练，也可以用作高校教学及研究的问答知识库、高校各专业领域的知识图谱、大数据的教育教学及研究等。

High-quality question-answer paired text corpora. The corpora are categorized into four major domains: arts, science and engineering, medicine, and general category, covering more than 20 professional fields including finance, healthcare, law, electric power, industry, machinery, maintenance, daily life, encyclopedia and others. The dataset includes two main forms: knowledge-based single-turn question-answer data and scenario-based continuous multi-turn question-answer data. It can be directly used for the research and development (R&D) and application of artificial intelligence-based natural language processing (NLP), as well as serving as training data for the development of large language models (LLMs) and AI-generated content (AIGC) applications. Additionally, it can function as a question-answer knowledge base for university teaching and research, knowledge graphs for various professional disciplines in colleges and universities, and support big data-enabled education, teaching and research activities.

搜集汇总

背景与挑战

背景概述

该数据集是一个高质量的问答文本语料库，涵盖文科、理工科、医学和综合四个类别，包含金融、医疗、法律等20余个专业领域的知识，提供知识性一问一答和场景性连续多轮问答两种形式。它主要用于人工智能自然语言处理研发、大语言模型训练以及高校教学研究，数据大小为500 MB，为静态更新，适用于科技创新场景。