Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Network Embedding via Deep Prediction Model

Xin Sun, Zenghui Song, Yongbo Yu, Junyu Dong, Claudia Plant, Christian Boehm

TL;DR

The paper introduces NEDP, a network embedding framework that combines a degree-weight biased random walk with a deep prediction model (RNN/LSTM) and Laplacian-based local-structure optimization to capture transfer behaviors and structural context in networks. It jointly preserves global context via sampling and local structure via LapEO, with a flexible embedding backbone in RNN/LSTM and edge representations for downstream tasks. The authors demonstrate strong, cross-task performance on diverse datasets for clustering, visualization, classification, reconstruction, and link prediction, often outperforming state-of-the-art baselines. This approach provides robust node and edge representations suitable for a range of network mining tasks and offers a foundation for further work on dynamic network representation learning.

Abstract

Network-structured data becomes ubiquitous in daily life and is growing at a rapid pace. It presents great challenges to feature engineering due to the high non-linearity and sparsity of the data. The local and global structure of the real-world networks can be reflected by dynamical transfer behaviors among nodes. This paper proposes a network embedding framework to capture the transfer behaviors on structured networks via deep prediction models. We first design a degree-weight biased random walk model to capture the transfer behaviors on the network. Then a deep network embedding method is introduced to preserve the transfer possibilities among the nodes. A network structure embedding layer is added into conventional deep prediction models, including Long Short-Term Memory Network and Recurrent Neural Network, to utilize the sequence prediction ability. To keep the local network neighborhood, we further perform a Laplacian supervised space optimization on the embedding feature representations. Experimental studies are conducted on various datasets including social networks, citation networks, biomedical network, collaboration network and language network. The results show that the learned representations can be effectively used as features in a variety of tasks, such as clustering, visualization, classification, reconstruction and link prediction, and achieve promising performance compared with state-of-the-arts.

Network Embedding via Deep Prediction Model

TL;DR

The paper introduces NEDP, a network embedding framework that combines a degree-weight biased random walk with a deep prediction model (RNN/LSTM) and Laplacian-based local-structure optimization to capture transfer behaviors and structural context in networks. It jointly preserves global context via sampling and local structure via LapEO, with a flexible embedding backbone in RNN/LSTM and edge representations for downstream tasks. The authors demonstrate strong, cross-task performance on diverse datasets for clustering, visualization, classification, reconstruction, and link prediction, often outperforming state-of-the-art baselines. This approach provides robust node and edge representations suitable for a range of network mining tasks and offers a foundation for further work on dynamic network representation learning.

Abstract

Network-structured data becomes ubiquitous in daily life and is growing at a rapid pace. It presents great challenges to feature engineering due to the high non-linearity and sparsity of the data. The local and global structure of the real-world networks can be reflected by dynamical transfer behaviors among nodes. This paper proposes a network embedding framework to capture the transfer behaviors on structured networks via deep prediction models. We first design a degree-weight biased random walk model to capture the transfer behaviors on the network. Then a deep network embedding method is introduced to preserve the transfer possibilities among the nodes. A network structure embedding layer is added into conventional deep prediction models, including Long Short-Term Memory Network and Recurrent Neural Network, to utilize the sequence prediction ability. To keep the local network neighborhood, we further perform a Laplacian supervised space optimization on the embedding feature representations. Experimental studies are conducted on various datasets including social networks, citation networks, biomedical network, collaboration network and language network. The results show that the learned representations can be effectively used as features in a variety of tasks, such as clustering, visualization, classification, reconstruction and link prediction, and achieve promising performance compared with state-of-the-arts.

Paper Structure

This paper contains 23 sections, 15 equations, 10 figures, 11 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Problem Definition
  4. Model Description
  5. Degree-Weight biased Random Walk
  6. The Embedding Model
  7. Recurrent Neural Network
  8. Long Short-Term Memory Network
  9. Embedding Space Optimization
  10. Edge Representations
  11. Experiments
  12. Datasets
  13. Baseline Methods
  14. Parameter Settings
  15. Experiments Results
  16. ...and 8 more sections

Figures (10)

  • Figure 1: The proposed Network Embedding framework via Deep Prediction models (NEDP) framework. $M$ is an available modular which includes RNN and LSTM.
  • Figure 2: A toy example of subgraphs. Social role and relationship is quite different among nodes $i$, $j$ and $k$.
  • Figure 3: Recurrent neural network (RNN)
  • Figure 4: Long Short-Term Memory Network (LSTM)
  • Figure 5: The core component in LSTM
  • ...and 5 more figures