The field of Go game research is hampered by the absence of records and analytical tools. In recent years, the increasing number of professional competitions and the advent of AlphaZero-based algorithms have provided an excellent opportunity for analyzing human games on a large scale. In this article, we present the ProfessionAl Go annotation datasEt (PAGE), containing 98 525 games played by 2007 professional players and spans over 70 years. PAGE incorporates both game-level metadata and in-game statistics across two dimensions. The game-level metadata pertinent to games, players, and tournaments is annotated by consolidating data from numerous reliable sources. The comprehensive in-game statistics are generated from the KataGo engine. Beyond the preliminary analysis of the dataset, we propose sample tasks that highlight PAGE's potential applications in various research areas. To the best of our knowledge, PAGE is the first dataset with extensive annotation in the game of Go . This article is an extended version of (Gao, 2022), where we perform a more detailed description, analysis, and applications.

中文翻译：

---

专业的 Go 注释数据集

的领域围棋研究因缺乏记录和分析工具而受到阻碍。近年来，职业比赛的不断增多以及基于AlphaZero的算法的出现，为大规模分析人类游戏提供了绝佳的机会。在这篇文章中，我们介绍了专业围棋注释数据集 (PAGE)，包含 2007 名职业棋手玩的 98 525 场比赛，时间跨度超过 70 年。 PAGE 结合了游戏级元数据和游戏内的二维统计数据。与游戏、玩家和锦标赛相关的游戏级元数据是通过整合来自众多可靠来源的数据来注释的。全面的游戏统计数据由 KataGo 引擎生成。除了对数据集的初步分析之外，我们还提出了示例任务，突出了 PAGE 在各个研究领域的潜在应用。据我们所知，PAGE 是游戏中第一个具有广泛注释的数据集去 。本文是（Gao，2022）的扩展版本，我们在其中进行了更详细的描述、分析和应用。