DerivedFunction/derivative-classification-multi-label

这是一个合成金融文本数据集，用于训练和评估自然语言处理模型在金融衍生品的多标签分类性能。数据集包含了使用预定义模板生成的文本，这些模板捕获了典型的金融披露模式，并用金融术语、公司名称、日期和数值数据填充占位符。数据集旨在解决金融领域中标记衍生品和对冲数据的稀缺问题，通过合成数据生成，实现了对标记句子的可控大规模创建，包括多标签注释。

This is a synthetic financial text dataset designed for training and evaluating NLP models on the multi-label classification of financial derivatives. The dataset consists of text generated using pre-defined templates that capture typical financial disclosure patterns, with placeholders filled in with financial terminology, company names, dates, and numerical data. The dataset aims to address the scarcity of labeled derivative and hedge data in the financial domain, enabling the controlled, large-scale creation of labeled sentences with multi-label annotations.