Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Deep Structural Knowledge Exploitation and Synergy for Estimating Node Importance Value on Heterogeneous Information Networks

Yankai Chen, Yixiang Fang, Qiongyan Wang, Xin Cao, Irwin King

TL;DR

A novel learning framework namely SKES is proposed, different from previous automatic learning designs, which exploits heterogeneous structural knowledge to enrich the informativeness of node representations and establishes an interpretable node importance computation paradigm.

Abstract

Node importance estimation problem has been studied conventionally with homogeneous network topology analysis. To deal with network heterogeneity, a few recent methods employ graph neural models to automatically learn diverse sources of information. However, the major concern revolves around that their full adaptive learning process may lead to insufficient information exploration, thereby formulating the problem as the isolated node value prediction with underperformance and less interpretability. In this work, we propose a novel learning framework: SKES. Different from previous automatic learning designs, SKES exploits heterogeneous structural knowledge to enrich the informativeness of node representations. Based on a sufficiently uninformative reference, SKES estimates the importance value for any input node, by quantifying its disparity against the reference. This establishes an interpretable node importance computation paradigm. Furthermore, SKES dives deep into the understanding that "nodes with similar characteristics are prone to have similar importance values" whilst guaranteeing that such informativeness disparity between any different nodes is orderly reflected by the embedding distance of their associated latent features. Extensive experiments on three widely-evaluated benchmarks demonstrate the performance superiority of SKES over several recent competing methods.

Deep Structural Knowledge Exploitation and Synergy for Estimating Node Importance Value on Heterogeneous Information Networks

TL;DR

A novel learning framework namely SKES is proposed, different from previous automatic learning designs, which exploits heterogeneous structural knowledge to enrich the informativeness of node representations and establishes an interpretable node importance computation paradigm.

Abstract

Node importance estimation problem has been studied conventionally with homogeneous network topology analysis. To deal with network heterogeneity, a few recent methods employ graph neural models to automatically learn diverse sources of information. However, the major concern revolves around that their full adaptive learning process may lead to insufficient information exploration, thereby formulating the problem as the isolated node value prediction with underperformance and less interpretability. In this work, we propose a novel learning framework: SKES. Different from previous automatic learning designs, SKES exploits heterogeneous structural knowledge to enrich the informativeness of node representations. Based on a sufficiently uninformative reference, SKES estimates the importance value for any input node, by quantifying its disparity against the reference. This establishes an interpretable node importance computation paradigm. Furthermore, SKES dives deep into the understanding that "nodes with similar characteristics are prone to have similar importance values" whilst guaranteeing that such informativeness disparity between any different nodes is orderly reflected by the embedding distance of their associated latent features. Extensive experiments on three widely-evaluated benchmarks demonstrate the performance superiority of SKES over several recent competing methods.
Paper Structure (22 sections, 18 equations, 4 figures, 6 tables)

This paper contains 22 sections, 18 equations, 4 figures, 6 tables.

Table of Contents

  1. Preliminaries and Problem Formulation
  2. SKES Methodology
  3. Overview
  4. Structural Priori Knowledge Exploitation
  5. 1) Metapath-induced Sub-network Construction.
  6. 2) Priori Centrality and Similarity Embedding.
  7. Synergetic Representation of Feature Distribution
  8. 1) Heterogeneous Knowledge Aggregation.
  9. 2) Empirical Representation of Feature Distribution.
  10. Node Importance Value Estimation
  11. Training Objective
  12. Experiments
  13. Experiment Setups
  14. 1) Benchmarks.
  15. 2) Evaluation Metrics.
  16. ...and 7 more sections

Figures (4)

  • Figure 1: An HIN example and SKES methodology.
  • Figure 2: The framework of our proposed model (best view in color).
  • Figure 3: (1) Decreasing contribution of structural knowledge; (2) curves of evaluation MAE (best view in color).
  • Figure 4: Absolute importance value gap (best view in color)