Rug Pull Dataset

Rug Pull Dataset是由南洋理工大学的研究团队构建的一个全面的数据集，旨在更准确地反映DeFi领域中Rug Pull现象的多样性和普遍性。该数据集包含2360个实例，覆盖了19种不同的Rug Pull根本原因，从而将覆盖率从20%提升至56%。数据集的构建过程涉及对现有数据集的评估、真实世界实例的收集以及实例的详细分析和注释。该数据集主要用于研究和开发更有效的Rug Pull检测工具，以保护投资者利益和增强DeFi项目的信任度。

A research team from Nanyang Technological University developed the Rug Pull Dataset, a comprehensive resource aimed at accurately capturing the diversity and prevalence of rug pull incidents within the Decentralized Finance (DeFi) ecosystem. This dataset comprises 2,360 instances spanning 19 distinct root causes of rug pulls, thereby increasing the coverage rate from 20% to 56%. The construction workflow of the dataset includes the assessment of existing datasets, the collection of real-world cases, as well as detailed analysis and manual annotation of each collected instance. This dataset is primarily intended to support research and the development of more effective rug pull detection tools, with the goals of protecting investor interests and enhancing the trustworthiness of DeFi projects.