Overlapping community detection at scale: a nonnegative matrix factorization approach

By Jaewon Yang, Jure Leskovec. 2013

981 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Statistical and Nonlinear Physics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Yuanzhi Li
2. Shirui Pan
3. Yingyu Liang
4. Darko Hric
5. Rok Sosic
6. Bryan Perozzi
7. Santo Fortunato
8. Jure Leskovec

In Statistical and Nonlinear Physics

1. Darko Hric
2. Jure Leskovec
3. Philip S. Yu
4. Shirui Pan
5. Rok Sosic
6. Bryan Perozzi
7. Santo Fortunato
8. Guodong Long

Citations in the corpus

• Debiasing Community Detection: The Importance of Lowly Connected Nodes (2019, methodology): “Note that any community detection task that only utilizes the network, such as BigCLAM and DEMON [4], [5], can replace this step, which makes our algorithm flexible, in general, and also capable of handling overlapping communities.”
• Community detection in networks: A user guide (2016, methodology): “Over the past few years this has finally become possible, due to the availability of several large network data sets with annotated vertices (Hric et al., 2014; Yang and Leskovec, 2012a, 2013, 2014).”
• Dropout Training of Matrix Factorization and Autoencoder for Link Prediction in Sparse Graphs (2015, background): “Among the large number of models proposed over the decade, Matrix Factorization (MF) is one of the most popular ones in network modeling and relational learning in general [27, 24, 14, 22].”
• Scalable and Interpretable Product Recommendations via Overlapping Co-Clustering (2016, methodology): “This idea is taken from [48], where it was used in the context of community detection.”
• SNAP: A General Purpose Network Analysis and Graph Mining Library (2016, methodology): “SNAP provides efficient implementations of commonly used traditional algorithms for graph and network analysis, as well as recent algorithms that employ machine learning techniques on graph problems, such as community detection [Yang and Leskovec 2013; Yang and Leskovec 2014 III.”
• Local Modeling of Attributed Graphs: Algorithms and Applications (2016, methodology): “METIS [68], Spectral [109], Co-clustering [38] X X Autopart, Cross-associations [26] X X X PageRank-Nibble [8], [145], BigClam [152] X X X Spectral Relational [82], SA-Cluster [154] X X CoPaM [98], Gamer [57] X X X X PICS [3] X X X CODA [51] X X X GOutRank [101], ConSub [66] X X X X FOCUSCO [this pa”
• A Survey of Community Detection Approaches: From Statistical Modeling to Deep Learning (2021, background): “BIGCLAM (2013) [77] Present a cluster affiliation model for overlapping, hierarchically nested community detection in large scale networks.”
• MGAE: Marginalized Graph Autoencoder for Graph Clustering (2017, background): “Recently, over-lapping community detection algorithms, like BigClam [39] and AgmFit [38], have also been developed; however, these algorithms are suboptimal because they only use one channel of information and ignore the other.”
• Provable Alternating Gradient Descent for Non-negative Matrix Factorization with Strong Correlations (2017, positive): “Non-negative matrix factorization (NMF) is an important tool in data analysis and is widely used in image processing, text mining, and hyperspectral imaging (e.g., (Lee & Seung, 1997; Blei et al., 2003; Yang & Leskovec, 2013)).”