Duong, Chi ThangNguyen, Thanh TamHoang, Trung-DungYin, HongzhiWeidlich, MatthiasNguyen, Quoc Viet Hung2022-12-052022-12-052022-12-052022-02-0110.1016/j.patcog.2022.109126https://infoscience.epfl.ch/handle/20.500.14299/193030WOS:000882982300003We present Deep MinCut (DMC), an unsupervised approach to learn node embeddings for graph -structured data. It derives node representations based on their membership in communities. As such, the embeddings directly provide insights into the graph structure, so that a separate clustering step is no longer needed. DMC learns both, node embeddings and communities, simultaneously by minimizing the mincut loss , which captures the number of connections between communities. Striving for high scalabil-ity, we also propose a training process for DMC based on minibatches. We provide empirical evidence that the communities learned by DMC are meaningful and that the node embeddings are competitive in different node classification benchmarks. (c) 2022 Elsevier Ltd. All rights reserved.Computer Science, Artificial IntelligenceEngineering, Electrical & ElectronicComputer ScienceEngineeringnode embeddinggraph representation learningcommunity detectioninterpretable machine learningtext::journal::journal article::research article