hh123wa/data_sample_1000

TAAC2026演示数据集（1000个样本）是一个包含1000条用户-项目交互记录的样本数据集，专为TAAC2026竞赛设计。该数据集采用扁平列布局，所有特征都作为独立的顶级列存储，而非嵌套结构或数组。数据集包含120列，分为6个类别：ID与标签（5列，包括用户ID、项目ID、标签类型、标签时间和时间戳，均为int64或int32类型，无空值）、用户整数特征（46列，包括35列标量int64和11列列表int64）、用户密集特征（10列，均为列表float类型）、项目整数特征（14列，包括13列标量int64和1列列表int64）、领域序列特征（45列，均为列表int64类型，来自4个行为领域：领域A（9列）、领域B（14列）、领域C（12列）、领域D（10列））。文件为demo_1000.parquet，大小约39 MB。数据集已更新至最新版本，特征名称和部分特征有所调整。A sample dataset containing 1000 user-item interaction records for the TAAC2026 competition. This dataset uses a flat column layout — all features are stored as individual top-level columns instead of nested structs/arrays. It consists of 120 columns categorized into 6 groups: ID & Label (5 columns, including user_id, item_id, label_type, label_time, and timestamp, all int64/int32 with no null values), User Int Features (46 columns, with 35 scalar int64 and 11 list<int64>), User Dense Features (10 columns, all list<float>), Item Int Features (14 columns, with 13 scalar int64 and 1 list<int64>), and Domain Sequence Features (45 columns, all list<int64> from 4 behavioral domains: domain_a_seq (9 columns), domain_b_seq (14 columns), domain_c_seq (12 columns), domain_d_seq (10 columns)). The file is demo_1000.parquet, approximately 39 MB in size, and has been updated to the latest version with renamed features and updates.