49,168 Hits in 3.7 sec

Node Embedding over Temporal Graphs [article]

Uriel Singer and Ido Guy and Kira Radinsky
2019 arXiv   pre-print
In this work, we present a method for node embedding in temporal graphs.  ...  We propose an algorithm that learns the evolution of a temporal graph's nodes and edges over time and incorporates this dynamics in a temporal node embedding framework for different graph prediction tasks  ...  Other temporal representations include the application of generalized singular value decomposition to consecutive time steps [57] and the use of node attribute Node Embedding over Temporal Graphs , ,  ... 
arXiv:1903.08889v2 fatcat:ooelumbkmrdldo2dy563vpauga

A Survey on Embedding Dynamic Graphs [article]

Claudio D. T. Barros, Matheus R. F. Mendonça, Alex B. Vieira, Artur Ziviani
2021 arXiv   pre-print
Therefore, several methods for embedding dynamic graphs have been proposed to learn network representations over time, facing novel challenges, such as time-domain modeling, temporal features to be captured  ...  Embedding static graphs in low-dimensional vector spaces plays a key role in network analytics and inference, supporting applications like node classification, link prediction, and graph visualization.  ...  node embedding, edge embedding, substructure embedding, and graph snapshot embedding, all of them over time, and (ii) temporal embedding, regarding the relation between network temporal granularity given  ... 
arXiv:2101.01229v2 fatcat:lqjkkksn45g7beizhcstakf6ry

DynG2G: An Efficient Stochastic Graph Embedding Method for Temporal Graphs [article]

Mengjia Xu, Apoorva Vikram Singh, George Em Karniadakis
2022 arXiv   pre-print
However, recent advances mostly focus on learning node embeddings as deterministic "vectors" for static graphs yet disregarding the key graph temporal dynamics and the evolving uncertainties associated  ...  Dynamic graph embedding has gained great attention recently due to its capability of learning low dimensional graph representations for complex temporal graphs with high accuracy.  ...  Uncertainty quantification for node embeddings The existing temporal graph embedding methods are deterministic and they transform the temporal graph nodes as fixed points in the latent space, hence ignoring  ... 
arXiv:2109.13441v2 fatcat:klz5o2q6uvbjrprzlh2t46retq

From Static to Dynamic Node Embeddings [article]

Di Jin, Sungchul Kim, Ryan A. Rossi, Danai Koutra
2020 arXiv   pre-print
the notion of an ϵ-graph time-series that uses a fixed number of edges for each graph, and show its superiority over the time-scale representation used in previous work.  ...  Furthermore, the dynamic embedding methods from the framework almost always achieve better predictive performance than existing state-of-the-art dynamic node embedding methods that are developed specifically  ...  To combine the embeddings over the graph time-series, we follow Algorithm 2. Temporal Embeddings 4.3.1 Base embedding methods.  ... 
arXiv:2009.10017v1 fatcat:ri2odfqmafavbghx2y4y4pa5ja

Dynamic Graph Embedding via LSTM History Tracking [article]

Shima Khoshraftar, Sedigheh Mahdavi, Aijun An, Yonggang Hu, Junfeng Liu
2019 arXiv   pre-print
In this paper, we present a dynamic network embedding method that integrates the history of nodes over time into the current state of nodes.  ...  the time and memory by training an autoencoder LSTM model using temporal walks rather than adjacency matrices of graphs which are the common practice.  ...  Dynamic graphs can change with different rates over time and we need mechanisms to reflect temporal changes in the node embeddings.  ... 
arXiv:1911.01551v1 fatcat:355zo5hsgvff5odkwnhwg5ypny

GRADE: Graph Dynamic Embedding [article]

Simeon Spasov, Alessandro Di Stefano, Pietro Lio, Jian Tang
2021 arXiv   pre-print
At each time step link generation is performed by first assigning node membership from a distribution over the communities, and then sampling a neighbor from a distribution over the nodes for the assigned  ...  Existing approaches for modelling graph dynamics focus extensively on the evolution of individual nodes independently of the evolution of mesoscale community structures.  ...  Introducing temporal evolution Random walks Fig. 1 : In our dynamic graph model, we assume that the semantic meaning of communities and the social context of individual nodes evolve over time.  ... 
arXiv:2007.08060v3 fatcat:ceiicjx5wvctbklgfbizirrnqq

Spatio-Temporal Deep Graph Infomax [article]

Felix L. Opolka, Aaron Solomon, Cătălina Cangea, Petar Veličković, Pietro Liò, R Devon Hjelm
2019 arXiv   pre-print
Spatio-temporal graphs such as traffic networks or gene regulatory systems present challenges for the existing deep learning methods due to the complexity of structural changes over time.  ...  Our model tackles the challenging task of node-level regression by training embeddings to maximize the mutual information between patches of the graph, at any given time step, and between features of the  ...  CONCLUSION We have presented STDGI, an approach for learning embeddings of nodes in a graph that evolves over time, which leverages mutual information maximization and performs node regression in a spatio-temporal  ... 
arXiv:1904.06316v1 fatcat:g3mkcjy5bjccpeymw5i6vo56ii

Dynamic Node Embeddings from Edge Streams [article]

John Boaz Lee, Giang Nguyen, Ryan A. Rossi, Nesreen K. Ahmed, Eunyee Koh, Sungchul Kim
2020 arXiv   pre-print
Networks evolve continuously over time with the addition, deletion, and changing of links and nodes.  ...  Unlike previous work, CTDNEs learn dynamic node embeddings directly from the temporal network at the finest temporal granularity and thus use only temporally valid information.  ...  Hence, the context and resulting embedding of a node changes temporally as the graph evolves over time.  ... 
arXiv:1904.06449v2 fatcat:nmfuo63f2zdbfkf6v7on235gsu

Temporal network embedding using graph attention network

Anuraj Mohan, K V Pramod
2021 Complex & Intelligent Systems  
First, we perform a temporal walk over the network to generate a positive pointwise mutual information matrix (PPMI) which denote the temporal correlation between the nodes.  ...  In this work, we propose a temporal graph attention network (TempGAN), where the aim is to learn representations from continuous-time temporal network by preserving the temporal proximity between nodes  ...  We aim to extend the concept of graph convolution and attention to temporal network data so as to generate time-aware node embeddings.  ... 
doi:10.1007/s40747-021-00332-x fatcat:jy6q2meccnbqvjhhxkdlmmhnmm

Dynamic Graph Representation Learning via Self-Attention Networks [article]

Aravind Sankar, Yanhong Wu, Liang Gou, Wei Zhang, Hao Yang
2019 arXiv   pre-print
Our experimental results show that DySAT has a significant performance gain over several different state-of-the-art graph embedding baselines.  ...  Previous methods on graph representation learning mainly focus on static graphs, however, many real-world graphs are dynamic and evolve over time.  ...  Unlike static graph embedding methods that focus entirely on preserving structural proximity, we learn dynamic node representations that reflect the temporal evolution of graph structure over a varying  ... 
arXiv:1812.09430v2 fatcat:qfilyfvmj5ecjmiknkvoropeqi

Representation Learning over Dynamic Graphs [article]

Rakshit Trivedi, Mehrdad Farajtabar, Prasenjeet Biswal, Hongyuan Zha
2018 arXiv   pre-print
The learned embeddings drive the dynamics of two key processes namely, communication and association between nodes in dynamic graphs.  ...  How can we effectively encode evolving information over dynamic graphs into low-dimensional representations?  ...  learn graph embeddings over static graphs with a smart random walk based approach to select the neighborhood to learn individual node embeddings.  ... 
arXiv:1803.04051v2 fatcat:gcgfprjh6fhpbhgcegbou4aiee

Dynamic Heterogeneous Graph Embedding Using Hierarchical Attentions [chapter]

Luwei Yang, Zhibo Xiao, Wen Jiang, Yi Wei, Yi Hu, Hao Wang
2020 Lecture Notes in Computer Science  
In this paper, we propose a novel dynamic heterogeneous graph embedding method using hierarchical attentions (DyHAN) that learns node embeddings leveraging both structural heterogeneity and temporal evolution  ...  There is still a lack of research on dynamic heterogeneous graph embedding.  ...  To compute the final node embedding, we use h T i to attend over all its historically-temporal representations, {h 1 i , h 2 i , ..., h T −1 i }.  ... 
doi:10.1007/978-3-030-45442-5_53 fatcat:euq27o7eh5f6xaejnuqhnxpuda

Learning Dynamic Preference Structure Embedding From Temporal Networks [article]

Tongya Zheng, Zunlei Feng, Yu Wang, Chengchao Shen, Mingli Song, Xingen Wang, Xinyu Wang, Chun Chen, Hao Xu
2021 arXiv   pre-print
In the second stage, an additional attention layer is designed to fuse two sampled temporal subgraphs of a node, generating temporal node embeddings for downstream tasks.  ...  The dynamics of temporal networks lie in the continuous interactions between nodes, which exhibit the dynamic node preferences with time elapsing.  ...  devise an attention-based fusion layer to fuse node embeddings from different temporal subgraphs sampled by TDS and GAS. • Experiments are conducted over a wide range of temporal networks, and results  ... 
arXiv:2111.11886v1 fatcat:iq2mdgfsf5a7resoakpm22w6v4

ConvDySAT: Deep Neural Representation Learning on Dynamic Graphs via Self-Attention and Convolutional Neural Networks [article]

Ahmad Hafez, Atulya Praphul, Yousef Jaradt, Ezani Godwin
2021 arXiv   pre-print
Learning node representations on temporal graphs is a fundamental step to learn real-word dynamic graphs efficiently.  ...  Real-world graphs have the nature of continuously evolving over time, such as changing edges weights, removing and adding nodes and appearing and disappearing of edges, while previous graph representation  ...  ., 2018 extended the idea to graphs by enabling the nodes in the graph to attend over the neighbouring nodes, outperforming most of the state-ofthe-art static graph embedding methods.  ... 
arXiv:2106.11430v1 fatcat:t2r2oycql5gttk3z22qblkrk3q

Modeling Event Propagation via Graph Biased Temporal Point Process [article]

Weichang Wu, Huanxi Liu, Xiaohu Zhang, Yu Liu, Hongyuan Zha
2020 arXiv   pre-print
Moreover, the learned node embedding vector is also integrated into the embedded event history as side information.  ...  We propose a Graph Biased Temporal Point Process (GBTPP) leveraging the structural information from graph representation learning, where the direct influence between nodes and indirect influence from event  ...  node embedding vector is obtained by reducing the historical T embeddings of each node into a final one via RNN, so that the evolution of a temporal graph over time can be captured.  ... 
arXiv:1908.01623v2 fatcat:aex4vt4d7jg7hii74r6hrj3guy
« Previous Showing results 1 — 15 out of 49,168 results