Galaxy Network Embedding: A Hierarchical Community Structure Preserving Approach

Du, Lun; Lu, Zhicong; Wang, Yun; Song, Guojie; Wang, Yiming; Chen, Wei

doi:10.24963/ijcai.2018/287

Cited by 53 publications

(36 citation statements)

References 10 publications

(1 reference statement)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Definition 3.3. Member Similarity is the ratio of the common members to all members in two ground-truth communities, which is extended from the definition of common neighbor similarity [11]:…”

Section: Definition 32mentioning

confidence: 99%

Tag2Vec: Learning Tag Representations in Tag Networks

Wang

Song

et al. 2019

The World Wide Web Conference

Self Cite

View full text Add to dashboard Cite

Network embedding is a method to learn low-dimensional representation vectors for nodes in complex networks. In real networks, nodes may have multiple tags but existing methods ignore the abundant semantic and hierarchical information of tags. This information is useful to many network applications and usually very stable. In this paper, we propose a tag representation learning model, Tag2Vec, which mixes nodes and tags into a hybrid network. Firstly, for tag networks, we define semantic distance as the proximity between tags and design a novel strategy, parameterized random walk, to generate context with semantic and hierarchical information of tags adaptively. Then, we propose hyperbolic Skip-gram model to express the complex hierarchical structure better with lower output dimensions. We evaluate our model on the NBER U.S. patent dataset and WordNet dataset. The results show that our model can learn tag representations with rich semantic information and it outperforms other baselines.

show abstract

Section: Definition 32mentioning

confidence: 99%

Tag2Vec: Learning Tag Representations in Tag Networks

Wang

Song

et al. 2019

The World Wide Web Conference

Self Cite

View full text Add to dashboard Cite

show abstract

“…Since hierarchy greatly enriches the network structure by introducing the concept of community, related researches on network embedding are devoted to mining this relation and profits a lot from it. MNMF [14] preserves community structures by nonnegative matrix factorization and GNE [15] introduces spherical projection for encoding hierarchical structures. However, these methods are based on transductive learning with the limitation that it is hard to generalize to new nodes.…”

Section: Introductionsmentioning

confidence: 99%

Fast Hierarchy Preserving Graph Embedding via Subspace Constraints

Chen

Wang

et al. 2021

ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

Hierarchy preserving network embedding is a method that project nodes into feature space by preserving the hierarchy property of networks. Recently, researches on network representation have considerably profited from taking hierarchy into consideration. Among these works, SpaceNE 1 [1] stands out by preserving hierarchy with the help of subspace constraints on the hierarchy subspace system. However, like all other hierarchy preserving network embedding methods, SpaceNE is time-consuming and cannot generalize to new nodes. In this paper, we propose an inductive method, FastHGE, to learn node representations more efficiently and generalize to new nodes more easily. Empirically, the experiment of node classification demonstrates that the convergence speed of FastHGE is increased by 30 times in the case of the same accuracy with SpaceNE.

show abstract

“…In almost all cases, definitions and identification methods have been described as simple, incomplete, reliant on intuition, and based on an assumption that anything that is not weak must be strong. Researchers have generally failed to take into account overlapping and hierarchical community properties that accurately reflect the organizational structures and functional components of complex networks commonly found in real-world domains [2,[36][37][38]. Overlaps indicate that some nodes belong to more than one community, and hierarchies indicate a possibility of creating numerous small-scale communities with the potential to aggregate into larger ones.…”

Section: Introductionmentioning

confidence: 99%

Distinguishing Arc Types to Understand Complex Network Strength Structures and Hierarchical Connectivity Patterns

Huang

Chin

2020

IEEE Access

View full text Add to dashboard Cite

Many real-world networks consisting of nodes representing (in)tangible asymmetric information or energy flows must be modeled as directed graphs (digraphs). Several methods for classifying non-directional edges in terms of strong or weak ties have been developed for well-known non-directional networks, but none specifically for directed networks. In almost all cases, definitions and identification methods are simple, incomplete, reliant on intuition, and based on the assumption that anything that is not weak must be strong. Researchers have generally failed to consider overlapping and hierarchical community properties that accurately reflect organizational structures or the functional components commonly found in real-world complex networks, resulting in multiple challenges to analyzing many types of directed networks. In this paper we describe a method that considers asymmetric definitions of arc strength, especially when arcs hold important directional significance. To more fully capture overlapping and hierarchical network community structures, we used hierarchy-based definitions to identify bond arcs, kth-layer local bridges, global bridges, and silk arcs and to create a hierarchical arc type analysis (HATA) algorithm. The algorithm employs a mix of common middle node measures and statistical parameters generated by randomized directed networks corresponding to the network being investigated. To test the HATA algorithm, we conducted four experiments involving a mix of arc rewiring and additions, multiple datasets associated with the Travian game, 56 empirical networks from previous studies, and 3 bird song transition networks. Our results indicate that HATA offers a novel perspective to understanding arc strengths and structures in directed complex networks. INDEX TERMS Bond arcs, common middle node, digraphs, directed arcs, global bridges, kth-layer local bridges, silk arcs.

show abstract

Galaxy Network Embedding: A Hierarchical Community Structure Preserving Approach

Cited by 53 publications

References 10 publications

Tag2Vec: Learning Tag Representations in Tag Networks

Tag2Vec: Learning Tag Representations in Tag Networks

Fast Hierarchy Preserving Graph Embedding via Subspace Constraints

Distinguishing Arc Types to Understand Complex Network Strength Structures and Hierarchical Connectivity Patterns

Contact Info

Product

Resources

About