FinS-Pilot

该数据集是一个用于评估在线金融应用中RAG系统的基准，它由真实世界的金融助手互动构建而成，融合了实时API数据及结构化文本资源。该基准通过一个意图分类框架进行组织，涵盖了诸如股票分析及宏观经济预测等关键金融领域。其任务是对金融助手在处理静态知识及时效性市场信息方面的能力进行评估。

This dataset is a benchmark for evaluating Retrieval-Augmented Generation (RAG) systems in online financial applications. It is constructed from real-world financial assistant interactions, integrating real-time API data and structured textual resources. This benchmark is organized via an intent classification framework, covering key financial domains such as stock analysis and macroeconomic forecasting. Its core task is to assess the capabilities of financial assistants in handling static knowledge and time-sensitive market information.