Abstract
We focus on the problem of predicting social media user’s future behavior and consider it as a graph node binary classification task. Existing works use graph representation learning methods to give each node an embedding vector, then update the node representations by designing different information passing and aggregation mechanisms, like GCN or GAT methods. In this paper, we follow the fact that social media users have influence on their neighbor area, and extract subgraph structures from real-world social networks. We propose an encoder–decoder architecture based on graph U-Net, known as the graph U-Net+. In order to improve the feature extraction capability in convolutional process and eliminate the effect of over-smoothing problem, we introduce the bilinear information aggregator and NodeNorm normalization approaches into both encoding and decoding blocks. We reuse four datasets from DeepInf and extensive experimental results demonstrate that our methods achieve better performance than previous models.
Highlights
With the explosive appearance of social media such as Twitter, WeChat, and Weibo, many people have joined the Internet and formed some huge social network neighborhoods
Our proposed model is composed of GCNII and bilinear information aggregator module without involving more advanced graph convolution layers like GAT, but AUC ad F1 metrics on four datasets show that graph U-Net+ performs better than graph self-attention based DeepInf-GAT
As for the less performance of precision and recall metrics on OAG dataset, we believe that the reason why graph U-Net+ model proposed in this paper performs poorly on OAG dataset may be that the other three datasets are all standard social networks based on the real world media, in which user interaction and friend relationship are more close to the research interest of this paper
Summary
With the explosive appearance of social media such as Twitter, WeChat, and Weibo, many people have joined the Internet and formed some huge social network neighborhoods. These social media allow users to directly view the contents of all their online friends, and support users to further improve the breadth and depth of information propagation through online interactions, such as subscribing, sharing, and retweeting. During the process of user interaction, user-level social influence naturally appears and imperceptively “interferes” with the behavior and judgment of each user [1,2,3]. We choose whether to watch movies based on suggestion of our closest friend. For another instance, a person may decide whether to believe a piece of political news based on what an authoritative official says. The resulting social user behavior prediction task has important value in many application fields, such as online advertising [4, 5], recommendation systems [6,7,8], viral influence prediction in marketing [9,10,11], rumor spreading [12], and even presidential elections [13]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
