loyoladatamining/usajobs

USAJOBS数据集是一个关于美国政府招聘职位发布的时间序列数据集，涵盖从2017年1月到2026年3月的月度数据。数据集包含两个配置：default和postings。default配置包括职位ID、文本内容、职位标题和公司/机构等特征；postings配置包括USAJOBS控制编号、文本内容和职位标题。数据集规模在100万到1000万条记录之间，总大小约84.28GB，下载大小约30.56GB。每个月度分割都提供了具体的示例数量和字节大小，例如2026年3月有14,145个示例。数据集适用于招聘分析、政府职位趋势研究、自然语言处理和时间序列分析等任务，语言为英语，许可证为MIT。The USAJOBS dataset is a time-series dataset of U.S. government job postings, covering monthly data from January 2017 to March 2026. The dataset includes two configurations: default and postings. The default configuration features job ID, text content, job title, and company/agency; the postings configuration features USAJOBS control number, text content, and job title. The dataset scale is between 1 million and 10 million records, with a total size of approximately 84.28GB and a download size of about 30.56GB. Each monthly split provides specific example counts and byte sizes, e.g., March 2026 has 14,145 examples. The dataset is suitable for recruitment analysis, government job trend research, natural language processing, and time-series analysis tasks. The language is English, and the license is MIT.