electricsheepeurope/europe-ilo-how-temp-eco-ocu-nb-mean-weekly-hours-actually-worked-per-employed-per
收藏资源简介:
该数据集包含来自国际劳工组织(ILO)ILOSTAT数据库的欧洲劳动力市场数据,专注于按经济活动和职业划分的就业人员实际平均每周工作时数这一指标。数据集涵盖38个欧洲国家从1991年到2025年的441,005个观测值,每年更新。数据通过ILOSTAT REST API获取,并经过 harmonization 处理,确保符合国际劳工统计学家会议(ICLS)的定义。数据集包含国家代码、国家名称、数据来源、指标代码、指标名称、经济活动和职业分类变量、观测年份、观测值、观测状态标志以及相关注释等字段。数据质量方面,该数据集为年度频率,当同一国家×年份有多个数据源时,采用ILO选择的最佳来源,且分类列仅在指标发布细分数据时才非空。该数据集适用于表格分类、回归和时间序列预测等机器学习任务,由Electric Sheep Europe重新打包为Parquet格式,便于使用Hugging Face的`load_dataset()`函数直接加载。
This dataset contains European labour market data from the International Labour Organization (ILO) ILOSTAT database, focusing on the indicator Mean weekly hours actually worked per employed person by economic activity and occupation. It includes 441,005 observations across 38 European countries from 1991 to 2025, updated annually. Data is pulled directly from the ILOSTAT REST API and harmonized using International Conference of Labour Statisticians (ICLS) definitions. The dataset features columns such as country code, country name, data source, indicator code, indicator name, economic activity and occupation classification variables, observation year, observed value, observation status flags, and relevant notes. Regarding data quality, the dataset is annual frequency; when multiple sources exist for the same country×year, the ILO-selected best source is used, and disaggregation columns are non-null only when the indicator publishes that breakdown. It is suitable for machine learning tasks like tabular classification, regression, and time-series forecasting, repackaged by Electric Sheep Europe into Parquet format for easy loading via Hugging Faces `load_dataset()` function.




