jc4p/farcaster-casts

Farcaster公共帖子数据集包含来自Farcaster社交协议的183,941,530条未被作者删除的公共帖子。数据集为每个帖子提供了详尽的元数据，包括用户ID、消息类型、时间戳、唯一哈希标识、签名方案、签名、签名者公钥、文本内容、提及的用户、父帖子ID、外部内容URL、嵌入内容以及目标帖子哈希等字段。数据集可用于分析Farcaster生态系统的对话模式、社交互动、回复链、社区形成、讨论和趋势的演变、社交媒体话语的语言模型训练、嵌入式内容共享模式分析、去中心化社交网络的增长和用户行为研究、创建帖子病毒传播可能性分类器以及基于互动构建Farcaster用户网络图等。

The Farcaster Public Casts Dataset contains 183,941,530 public posts from the Farcaster social protocol that have not been deleted by their authors. The dataset provides comprehensive metadata for each post, including fields such as user ID, message type, timestamp, unique hash identifier, cryptographic signature scheme, signature, signers public key, text content, mentioned users, parent post ID, external content URL, embedded content, and target post hash. The dataset can be used for analyzing conversation patterns and social interactions on Farcaster, studying reply chains and community formation, tracking the evolution of discussions and trends, training language models on social media discourse, analyzing embedded content sharing patterns, researching the growth of decentralized social networks and user behavior, creating classifiers of the likelihood of a post going viral, and building network graphs of Farcaster users based on interactions.