GCN-IA: User Profile Based on Graph Convolutional Network with Implicit Association Labels

Inferring multi-label user profile plays a significant role in providing individual recommendations and exact-marketing, etc. Current researches on multi-label user profile either ignore the implicit associations among labels or do not consider the user and label semantic information in the social networks. Therefore, the user profile inferred always does not take full advantage of the global information sufficiently. To solve above problem, a new insight is presented to introduce implicit association labels as the prior knowledge enhancement and jointly embed the user and label semantic information. In this paper, a graph convolutional network with implicit associations (GCN-IA) method is proposed to obtain user profile. Specifically, a probability matrix is first designed to capture the implicit associations among labels for user representation. Then, we learn user embedding and label embedding jointly based on user-generated texts, relationships and label information. On four real-world datasets in Weibo, experimental results demonstrate that GCN-IA produces a significant improvement compared with some state-of-the-art methods.