Abstract:Searching for a community containing a query node in an online social network enjoys wide applications like recommendation, team organization, etc. When applied to real-life networks, the existing approaches face two major limitations. First, they usually take two steps, i.e., crawling a large part of the network first and then finding the community next, but the entire network is usually too big and most of the data are not interesting to end users. Second, the existing methods utilize hand-crafted rules to measure community membership, while it is very difficult to define effective rules as the communities are flexible for different query nodes. This paper proposes an interactive community search method based on graph neural network (shortened by ICS-GNN(+)) to locate the target community over a subgraph collected on the fly from an online network iteratively. In each iteration, we first build a candidate subgraph around the query node and labeled nodes. We then train a node classification model using GNN to determine whether every node belongs to the target community, which captures similarities between nodes by combining content and structural features seamlessly and flexibly under the guide of users' labeling. Based on the probabilities inferred from the trained GNN, we introduce a k-sized Maximum-GNN-scores (shortened by kMG) community to describe the target community and design a method to locate the kMG community which will be evaluated by end users to acquire more feedback. Besides, various optimization strategies are proposed including an adaptive method to maintain the subgraph during iterations, combining ranking loss into the GNN model, generating node embedding enhanced by pseudo-labels from node clusters in the subgraph, and a greedy community searching method with benefit computed globally. We conduct the experiments on both offline and online real-life datasets, and demonstrate that ICS-GNN(+) can produce effective communities with low overhead in communication, computation, and user labeling.

CP-GNN: A Software for Community Detection in Heterogeneous Information Networks

Detecting User Community in Sparse Domain via Cross-Graph Pairwise Learning

Heterogeneous graph community detection method based on K-nearest neighbor graph neural network

FCS-HGNN: Flexible Multi-type Community Search in Heterogeneous Information Networks

Community Detection Method Based on GNN and Improved K-means

CommDGI

Gumbel-SoftMax based graph convolution network approach for community detection

CrimeGNN: Harnessing the Power of Graph Neural Networks for Community Detection in Criminal Networks

Towards a Better Tradeoff Between Quality and Efficiency of Community Detection: an Inductive Embedding Method Across Graphs

Deep graph convolution neural network with non-negative matrix factorization for community discovery

ICS-GNN+: Lightweight Interactive Community Search Via Graph Neural Network

Hypergraph network embedding for community detection

Hierarchical Message-Passing Graph Neural Networks

Supervised Community Detection with Line Graph Neural Networks

Directed Community Detection With Network Embedding

Edge classification based on Convolutional Neural Networks for community detection in complex network

HybridGNN: Learning Hybrid Representation in Multiplex Heterogeneous Networks

Region2Vec: Community Detection on Spatial Networks Using Graph Embedding with Node Attributes and Spatial Interactions

Graph Neural Networks with Heterophily

CommunityGAN: Community Detection with Generative Adversarial Nets

DCGNN: Adaptive deep graph convolution for heterophily graphs