Higher-order Network Representation Learning

@article{Rossi2018HigherorderNR,
  title={Higher-order Network Representation Learning},
  author={Ryan A. Rossi and Nesreen Ahmed and Eunyee Koh},
  journal={Companion Proceedings of the The Web Conference 2018},
  year={2018}
}
This paper describes a general framework for learning Higher-Order Network Embeddings (HONE) from graph data based on network motifs. The HONE framework is highly expressive and flexible with many interchangeable components. The experimental results demonstrate the effectiveness of learning higher-order network representations. In all cases, HONE outperforms recent embedding methods that are unable to capture higher-order structures with a mean relative gain in AUC of 19% (and up to 75% gain… 

Tables from this paper

HONEM: Network Embedding Using Higher-Order Patterns in Sequential Data

TLDR
It is demonstrated that the higher-order network embedding (HONEM) method is able to extract higher- order dependencies from HON to construct theHigher-order neighborhood matrix of the network, while existing methods are not able to capture these higher-orders.

HONEM: Learning Embedding for Higher Order Networks

TLDR
HONEM is a higher order network embedding method that captures the non-Markovian higher order dependencies in a network and outperforms other state-of-the-art methods in node classification, network reconstruction, link prediction, and visualization for networks that contain non- MarkovianHigher order dependencies.

Network Embedding via Motifs

TLDR
This work presents an algorithm for Learning Embeddings by leveraging Motifs Of Networks (LEMON), which aims to learn embeddings for vertices and various motifs, and finds that LEMON achieves significant improvements in downstream tasks.

Higher-Order Attribute-Enhancing Heterogeneous Graph Neural Networks

TLDR
Experimental results not only show HAEGNN’s superior performance against the state-of-the-art methods in node classification, node clustering, and visualization, but also demonstrate its superiorities in terms of memory efficiency and explainability.

When to Remember Where You Came from: Node Representation Learning in Higher-order Networks

TLDR
This work proposes a node representation learning framework called EVO or Embedding Variable Orders, which captures non-Markovian dependencies by combining work on higher-order networks with work on node embeddings, and shows that EVO outperforms baselines in tasks where high-order dependencies are likely to matter.

DynACPD Embedding Algorithm for Prediction Tasks in Dynamic Networks

TLDR
Novel embedding methods for a dynamic network based on higher order tensor decompositions for tensorial representations of the dynamic network where there exist both temporal and spatial link relationships between nodes.

OFFER: A Motif Dimensional Framework for Network Representation Learning

TLDR
By evaluating the performance of OFFER, both link prediction results and clustering results demonstrate that the graph representation learning algorithms enhanced with OFFER consistently outperform the original algorithms with higher efficiency.

Interactive Higher-Order Network Analysis

TLDR
This work develops interactive visual higher-order network mining and modeling techniques to gain insight into the higher-orders structure and composition of complex networks, including important phenotypes in a human gene interaction network and hubs in a power grid network.

A Directed Graph Link Prediction Method Combined with Higher Order Structure Information

TLDR
A link prediction model for directed networks called Motif-Based Graph Auto Encoder (MGAE) is proposed, and the proposed model is applied to solve the real-world problems on various types of networks, such as citation networks, web networks, and social networks.

Stacked Mixed-Order Graph Convolutional Networks for Collaborative Filtering

TLDR
This paper proposes SMOG-CF (Stacked Mixed-Order Graph Convolutional Networks for Collaborative Filtering), a GCN-based framework that can directly capture high-order connectivity among nodes and facilitates ‘path-level’ information propagation between neighboring nodes at any order.
...

References

SHOWING 1-10 OF 10 REFERENCES

GraRep: Learning Graph Representations with Global Structural Information

TLDR
A novel model for learning vertex representations of weighted graphs that integrates global structural information of the graph into the learning process and significantly outperforms other state-of-the-art methods in such tasks.

node2vec: Scalable Feature Learning for Networks

TLDR
In node2vec, an algorithmic framework for learning continuous feature representations for nodes in networks, a flexible notion of a node's network neighborhood is defined and a biased random walk procedure is designed, which efficiently explores diverse neighborhoods.

LINE: Large-scale Information Network Embedding

TLDR
A novel network embedding method called the ``LINE,'' which is suitable for arbitrary types of information networks: undirected, directed, and/or weighted, and optimizes a carefully designed objective function that preserves both the local and global network structures.

DeepWalk: online learning of social representations

TLDR
DeepWalk is an online learning algorithm which builds useful incremental results, and is trivially parallelizable, which make it suitable for a broad class of real world applications such as network classification, and anomaly detection.

Role Discovery in Networks

TLDR
A taxonomy of three general classes of techniques for discovering roles that includes (i) graph-based roles, (ii) feature- based roles, and (iii) hybrid roles is proposed, which consists of two fundamental components: (a) role feature construction and (b) role assignment using the learned feature representation.

Efficient Graphlet Counting for Large Networks

TLDR
This paper proposes a fast, efficient, and parallel algorithm for counting graphlets of size k={3,4}-nodes that take only a fraction of the time to compute when compared with the current methods used, and is on average 460x faster than current methods.

Biological network comparison using graphlet degree distribution

TLDR
Almost all of the 14 eukaryotic PPI networks, including human, resulting from various high-throughput experimental techniques, as well as from curated databases, are better modeled by geometric random graphs than by Erdös-Rény, random scale-free, or Barabási-Albert scale- free networks.

Leveraging social media networks for classification

TLDR
The proposed framework, SocioDim, first extracts social dimensions based on the network structure to accurately capture prominent interaction patterns between actors, then learns a discriminative classifier to select relevant social dimensions.

The Network Data Repository with Interactive Graph Analytics and Visualization

TLDR
The aim of NR is to make it easy to discover key insights into the data extremely fast with little effort while also providing a medium for users to share data, visualizations, and insights.

Edge Role Discovery via Higher-Order Structures

TLDR
This paper introduces edge role discovery and proposes a framework for learning and extracting edge roles from large graphs, and develops a novel edge feature learning approach for role discovery that begins with higher-order network motifs and automatically learns deeper edge features.