Heterogeneous Graph Structure Learning through the Lens of Data-generating Processes

2025-03-11

AI Summary

This work addresses Heterogeneous Graph Structure Learning (HGSL), aiming to automatically infer node- and edge-type topologies from observed data. We propose H²MN, the first statistical modeling framework for the Heterogeneous Graph Data Generating Process (DGP), formalizing HGSL as a Maximum A Posteriori (MAP) estimation problem grounded in the DGP. An alternating optimization algorithm with theoretical convergence guarantees is developed to solve it. Key contributions include: (1) the first interpretable and scalable heterogeneous graph DGP model; (2) the first unification of HGSL as a Bayesian graph inference problem; and (3) an efficient, robust structure learning solution. Extensive experiments on synthetic and real-world datasets demonstrate significant improvements in edge-type classification accuracy and edge-weight recovery fidelity, validating both effectiveness and generalizability.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Reasoning under Uncertainty: Graphical Models
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Web Mining and Content Analysis: Bridging structured and unstructured data

Abstract

Inferring the graph structure from observed data is a key task in graph machine learning to capture the intrinsic relationship between data entities. While significant advancements have been made in learning the structure of homogeneous graphs, many real-world graphs exhibit heterogeneous patterns where nodes and edges have multiple types. This paper fills this gap by introducing the first approach for heterogeneous graph structure learning (HGSL). To this end, we first propose a novel statistical model for the data-generating process (DGP) of heterogeneous graph data, namely hidden Markov networks for heterogeneous graphs (H2MN). Then we formalize HGSL as a maximum a-posterior estimation problem parameterized by such DGP and derive an alternating optimization method to obtain a solution together with a theoretical justification of the optimization conditions. Finally, we conduct extensive experiments on both synthetic and real-world datasets to demonstrate that our proposed method excels in learning structure on heterogeneous graphs in terms of edge type identification and edge weight recovery.

Problem

• Inferring graph structure from observed heterogeneous data.
• Proposing a novel statistical model for heterogeneous graph data.
• Developing an optimization method for heterogeneous graph structure learning.

Innovation

• Introduces heterogeneous graph structure learning (HGSL)
• Proposes hidden Markov networks for heterogeneous graphs (H2MN)
• Develops alternating optimization for maximum a-posterior estimation