polymarket-users
收藏资源简介:
Polymarket Users数据集是一个全面的链上预测市场研究数据集,涵盖了Polymarket(最大的链上预测市场)上所有用户的交易活动、利润和行为特征。数据集时间跨度为2022年11月11日至2026年3月29日,基于Polygon区块链上CTF Exchange合约的公开OrderFilled事件构建,并协调还原了最终用户交易、每日市值计价损益(PnL)以及丰富的用户级行为特征。数据集采用多表结构,包含元数据表(如市场元数据、事件元数据、预测代币映射)、用户特征与终端损益表(如user_features表包含约83列行为特征向量,user_pnl_summary表提供终端损益汇总)、每日损益面板(如pnl_daily及其变体)、损益变化面板(如每日和每月损益变化数据)以及交易级数据(如协调后的最终用户交易和聚合的OHLCV数据)。数据集适用于预测市场研究、用户行为分析、金融计量经济学和加密货币研究等任务。所有时间戳均采用UTC时间,并遵循特定的事件时间戳和快照时间戳约定。处理后的数据在CC-BY 4.0许可下发布,但源自Polymarket API的字段需遵守其使用条款。本数据集支持相关学术研究论文《Who Wins and Who Loses In Prediction Markets? Evidence from Polymarket》。
The Polymarket Users dataset is a comprehensive on-chain prediction market research dataset covering all user trading activities, profits, and behavioral characteristics on Polymarket, the largest on-chain prediction market. The dataset spans from November 11, 2022 to March 29, 2026, and is constructed based on public OrderFilled events from the CTF Exchange contract on the Polygon blockchain. It reconciles and restores final user trades, daily mark-to-market profit and loss (PnL), and rich user-level behavioral features. The dataset adopts a multi-table structure, including metadata tables (such as market metadata, event metadata, and prediction token mappings), user features and terminal profit and loss tables (e.g., the user_features table contains approximately 83 columns of behavioral feature vectors, and the user_pnl_summary table provides terminal profit and loss summaries), daily profit and loss panels (e.g., pnl_daily and its variants), profit and loss change panels (e.g., daily and monthly profit and loss change data), and trade-level data (such as reconciled final user trades and aggregated OHLCV data). The dataset is suitable for tasks such as prediction market research, user behavior analysis, financial econometrics, and cryptocurrency research. All timestamps are in UTC and follow specific event timestamp and snapshot timestamp conventions. The processed data is released under the CC-BY 4.0 license, but fields derived from the Polymarket API (such as market question text, descriptions, etc.) must comply with its terms of use. This dataset supports the related academic research paper Who Wins and Who Loses In Prediction Markets? Evidence from Polymarket.





