BMS-POS

该数据集名为BMS-POS，是一个关于产品购买交易的网络数据。其中，对象代表产品，属性代表购买交易，而边则表示产品在特定交易中被购买。本研究选取了该数据集中的前1946个交易记录。该数据集主要用于预测同一顾客可能购买的产品，并处理缺失的边信息。规模上，仅使用了原始数据集中的前1946个交易记录。研究任务包括对象-对象（O-O）和对象-属性（O-A）任务。

The dataset named BMS-POS is a network dataset concerning product purchase transactions. In this dataset, objects correspond to products, attributes correspond to purchase transactions, and edges represent the purchase of a product within a specific transaction. For this study, we selected the first 1946 transaction records from the original dataset. This dataset is primarily utilized for predicting products that a customer may potentially purchase and handling missing edge information. In terms of dataset scale, only the first 1946 transaction records from the original dataset were adopted. The research tasks include object-object (O-O) and object-attribute (O-A) tasks.