Abstract:Detecting the community structure within networks is important because it aids in the analysis of complex networks. However, the existence of diverse complex structures in the network topology can limit the topological information's ability to represent communities. For instance, the accuracy of many traditional methods based on topology alone decreases sharply when a link is missing or noisy. To solve the above-mentioned issues simultaneously, we propose a unified framework of semi-supervised community detection integrating content information. Three models are proposed, instead of a special model-handling problem. The framework used performs the same interpretation for a target problem. The inputs contain the adjacent and node content matrix. The proposed models describe the network topology using a non-negative matrix factorization (NMF) algorithm-based generative model or modularity maximization algorithm. Additionally, we use the equivalence theory of the pLSI, and NMF models the content information based on NMF. Community membership is also used as the relationship between communities and node contents. The models uniformly integrate content information and network topology. Further, a constraint matrix that depicts must-links is built by calculating the similarity of neighbors between any two nodes. Must-link pairs can represent the network topology's prior information. The models' community memberships are adjusted by using a strong constraint based on the connected component of this matrix. The models also introduce prior information into the topological and content information in the mentioned models. We compared the proposed models with the state-of-the-art community detection models on artificial networks and five real networks. The experimental results show that the proposed models outperformed the models combining topology and content and semi-supervised models. Fusing the topology with the content and the prior information in the models was also effective for community detection. Thus, the models under the framework are more advantageous than other semi-supervised models when incorporating the same amount of prior information into the models. The semi-supervised framework used also has a certain level of competitiveness in detecting communities.

Semisupervised Community Preserving Network Embedding with Pairwise Constraints.

Community Preserving Network Embedding

Network Embedding Using Semi-Supervised Kernel Nonnegative Matrix Factorization

Multi-Constrained Embedding for Accurate Community Detection on Undirected Networks

The Network Representation Learning Algorithm Based on Semi-Supervised Random Walk

Elsnc: A Semi-Supervised Community Detection Method with Integration of Embedding- Enhanced Links and Node Content in Attributed Networks

Learning Network Embedding with Community Structural Information

Lower Order Information Preserved Network Embedding Based on Non-Negative Matrix Decomposition

A Network Embedding-Enhanced NMF Method for Finding Communities in Attributed Networks.

Semi-supervised clustering via constrained symmetric non-negative matrix factorization

Attributed Network Embedding with Micro-Meso Structure

Sim2vec: Node Similarity Preserving Network Embedding.

Multi-Task Learning Based Network Embedding

Binarized network embedding with community structural information

Semi-supervised Symmetric Non-negative Matrix Factorization with Low-Rank Tensor Representation

Multi-Stage Network Embedding for Exploring Heterogeneous Edges

Network2Vec: Learning Node Representation Based on Space Mapping in Networks

Community Detection Method Based on Robust Semi-Supervised Nonnegative Matrix Factorization

Attributed Multi-layer Network Embedding.

Network Representation Learning Enhanced by Partial Community Information That Is Found Using Game Theory.

A Unified Framework of Semi-Supervised Community Detection Integrating Network Topology and Node Content