Learning multi-resolution representations of research patterns in bibliographic networks

•This study proposes a novel model for learning multi-resolution representations of bibliographic entities.•The proposed methods extract substructures of bibliographic networks and simplify the substructures according to multiple levels of detailedness.•The simplified subgraphs balance the learning opportunities of high- and low-performance bibliographic entities by co-occurring with both types of entities.•The proposed model discovered more consistent and multifaceted features of the bibliographic entities than the existing network embedding models.