新闻文本+道德判断
收藏资源简介:
## 数据集简介 该数据集包含基于道德基础分类的新闻文章,适用于中文的多标签文本分类任务。每篇新闻文章根据其道德倾向进行标注,分为实用(Pragmatism)/ 理想(Idealism)”、“责任(Responsibility)/ 利益(Profit)”、“创新(Innovation)/ 守旧(Conservatism)三个标签。 ## 许可证 该数据集的使用许可为 Apache License 2.0,您可以自由使用、修改和分发该数据集,但需遵循许可协议的规定。 ## 数据集格式 数据集以 CSV 格式提供,包含三个主要字段: - **system**:表示对模型的初始命令。 - **query**:表示新闻文章的文本内容。 - **response**:表示模型应返回的新闻文章的分类标签,支持多标签分类。 ## 任务和语言 - **任务类型**:文本分类(多标签) - **语言**:中文 ## 下载方法 您可以通过以下两种方式下载数据集: ### 使用 ModelScope SDK 下载 ```python from modelscope.msdatasets import MsDataset dataset = MsDataset.load('wowhaha/moral-foundation-news_2000')
## Dataset Overview This dataset contains news articles categorized based on moral foundations, designed for Chinese multi-label text classification tasks. Each news article is annotated based on its moral orientation, with three pairs of labels: "Pragmatism / Idealism", "Responsibility / Profit", and "Innovation / Conservatism". ## License This dataset is released under the Apache License 2.0. You are free to use, modify, and distribute the dataset, provided that you comply with the terms and conditions of the license agreement. ## Dataset Format The dataset is provided in CSV format, consisting of three core fields: - **system**: Denotes the initial prompt for the model. - **query**: Denotes the text content of the news article. - **response**: Denotes the classification labels that the model should return, supporting multi-label classification. ## Task and Language - **Task Type**: Text Classification (Multi-label) - **Language**: Chinese ## Download Methods You can download the dataset via the following two approaches: ### Download via ModelScope SDK python from modelscope.msdatasets import MsDataset dataset = MsDataset.load('wowhaha/moral-foundation-news_2000')




