Model training method, abnormal commodity identification method, electronic equipment and storage medium

The embodiment of the invention provides a model training method, an abnormal commodity identification method, electronic equipment and a storage medium. In the embodiment of the invention, graph neural network model training is carried out based on a risk graph reflecting a risk association relationship between sample commodities and an attribute graph reflecting an attribute association relationship between the sample commodities; the risk association relationship between the sample commodities and the attribute association relationship between the sample commodities are fully considered in the model training process, so that the target map neural network model obtained through training can effectively measure the risk correlation between the commodities, and then the abnormal commodities are accurately identified. 本申请实施例提供一种模型训练方法、异常商品识别方法、电子设备及存储介质。在本申请实施例中，基于反映样本商品之间的风险关联关系的风险图和反映样本商品之间的属性关联关系的属性图进行图神经网络模型训练，由于在模型训练过程中充分考虑样本商品之间的风险关联关系和样本商品之间的属性关联关系，这样训练得到的目标图神经网络模型能够有效地度量商品之间的风险相关性，进而准确识别异常商品。