ExioNAICS

该数据集是一个大规模的基准数据集，它由20,850家公司的文本描述汇编而成，这些描述都附有经过验证的北美行业分类系统（NAICS）标签，并与一个碳排放强度因子的经济模型相匹配。数据集分为80%的训练集、10%的验证集和10%的测试集，支持多个NAICS分类级别。该数据集涵盖了20,850家公司，涉及1,114个NAICS类别。其任务是基于NAICS标签进行碳排放估算与分类。

This is a large-scale benchmark dataset compiled from text descriptions of 20,850 companies. Each company's description is paired with verified North American Industry Classification System (NAICS) labels and matched to an economic model of carbon emission intensity factors. The dataset is split into 80% training set, 10% validation set and 10% test set, supporting multiple NAICS classification levels. It covers 20,850 companies spanning 1,114 NAICS categories, and its task is carbon emission estimation and classification based on NAICS labels.