Neural Embedding Propagation on Heterogeneous Networks

TL;DR

This paper introduces Neural Embedding Propagation (NEP), a novel semi-supervised learning method for heterogeneous networks that captures complex multi-typed interactions using distributed embeddings and modular networks, outperforming existing algorithms.

Contribution

NEP generalizes label propagation to heterogeneous networks, modeling complex interactions with dynamic modular networks and distributed embeddings, enabling effective semi-supervised learning on large-scale data.

Findings

NEP outperforms state-of-the-art algorithms in accuracy.

NEP is efficient and scalable to large networks.

NEP demonstrates robustness across multiple datasets.

Abstract

Classification is one of the most important problems in machine learning. To address label scarcity, semi-supervised learning (SSL) has been intensively studied over the past two decades, which mainly leverages data affinity modeled by networks. Label propagation (LP), however, as the most popular SSL technique, mostly only works on homogeneous networks with single-typed simple interactions. In this work, we focus on the more general and powerful heterogeneous networks, which accommodate multi-typed objects and links, and thus endure multi-typed complex interactions. Specifically, we propose \textit{neural embedding propagation} (NEP), which leverages distributed embeddings to represent objects and dynamically composed modular networks to model their complex interactions. While generalizing LP as a simple instance, NEP is far more powerful in its natural awareness of different types of objects and links, and the ability to automatically capture their important interaction patterns. Further, we develop a series of efficient training strategies for NEP, leading to its easy deployment on real-world heterogeneous networks with millions of objects. With extensive experiments on three datasets, we comprehensively demonstrate the effectiveness, efficiency, and robustness of NEP compared with state-of-the-art network embedding and SSL algorithms.