NodePiece: Compositional and Parameter-Efficient Representations of Large Knowledge Graphs
arXiv:2106.12144
Abstract
Conventional representation learning algorithms for knowledge graphs (KG) map each entity to a unique embedding vector. Such a shallow lookup results in a linear growth of memory consumption for storing the embedding matrix and incurs high computational costs when working with real-world KGs. Drawing parallels with subword tokenization commonly used in NLP, we explore the landscape of more parameter-efficient node embedding strategies with possibly sublinear memory requirements. To this end, we propose NodePiece, an anchor-based approach to learn a fixed-size entity vocabulary. In NodePiece, a vocabulary of subword/sub-entity units is constructed from anchor nodes in a graph with known relation types. Given such a fixed-size vocabulary, it is possible to bootstrap an encoding and embedding for any entity, including those unseen during training. Experiments show that NodePiece performs competitively in node classification, link prediction, and relation prediction tasks while retaining less than 10% of explicit nodes in a graph as anchors and often having 10x fewer parameters. To this end, we show that a NodePiece-enabled model outperforms existing shallow models on a large OGB WikiKG 2 graph having 70x fewer parameters.
arXiv commentAccepted to ICLR 2022
References in corpus (13)
- Distributed Representations of Words and Phrases and their Compositionality
- PyTorch: An Imperative Style, High-Performance Deep Learning Library
- A Survey on Knowledge Graphs: Representation, Acquisition and Applications
- A Review of Relational Machine Learning for Knowledge Graphs
- Pre-trained Models for Natural Language Processing: A Survey
- Scaling Laws for Neural Language Models
- Fast Graph Representation Learning with PyTorch Geometric
- RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space
- PyTorch-BigGraph: A Large-scale Graph Embedding System
- Bringing Light Into the Dark: A Large-scale Evaluation of Knowledge Graph Embedding Models Under a Unified Framework
- Knowledge Transfer for Out-of-Knowledge-Base Entities: A Graph Neural Network Approach
- Janossy Pooling: Learning Deep Permutation-Invariant Functions for Variable-Size Inputs
- SOLAR: Sparse Orthogonal Learned and Random Embeddings