FinChina SA
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/YerayL/FinChina-SA
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是一个专注于金融领域的企业早期预警的实体级细粒度情感分析数据集，包含了11,036篇新闻文章，并对其中21,272家公司的情感进行了标注。该数据集将情感分为正面和负面，而负面情感进一步细分为极度负面、非常负面和负面，以支持细粒度金融情感分析任务的开展。
This dataset is an entity-level fine-grained sentiment analysis dataset focused on enterprise early warning in the financial domain. It contains 11,036 news articles, with sentiment annotations for 21,272 companies covered in these articles. This dataset classifies sentiments into two broad categories: positive and negative, where the negative category is further subdivided into three sub-categories: extremely negative, highly negative, and general negative, to support the implementation of fine-grained financial sentiment analysis tasks.
