Fair perception social contact recommendation method and related device

The invention discloses a fair perception social contact recommendation method and a related device, and the method comprises the steps: building a heterogeneous social network based on node interaction data in the social network, carrying out the depolarization of user node features and structural semantic features containing sensitive attributes in the heterogeneous social network, carrying out the feature splicing after depolarization, employing a graph neural network, and carrying out the recommendation of the user node features and the structural semantic features. According to the method, the final features of all the nodes after depolarization are obtained, so that recommended items of each user node are determined, fair recommendation oriented to heterogeneous social networks is realized, and the problem of recommendation discrimination can be effectively solved. 本发明公开了一种公平感知社交推荐方法及相关装置，本发明基于社交网络中节点交互数据构建异质社交网络，对异质社交网络中的包含敏感属性的用户节点特征和结构语义特征去偏，并进行去偏后特征拼接，采用图神经网络，获得所有节点的去偏后的最终特征，从而确定每个用户节点的推荐物品，实现面向异质社交