Towards Continuous Reuse of Graph Models via Holistic Memory Diversification

2024-06-11

AI Summary

This work addresses catastrophic forgetting of prior-task knowledge during continual learning on dynamically growing graphs. Existing approaches neglect memory diversity and struggle to efficiently preserve broad prior knowledge within sparse graph memory buffers. To tackle these limitations, we propose the first holistic, diversity-aware memory selection and generation framework. Specifically, we introduce a greedy sampling strategy that jointly optimizes intra-class and inter-class diversity for buffer construction, and integrate variational graph embedding generation with adversarial variational learning to jointly enhance the completeness and generalizability of synthesized node representations. Extensive experiments on multiple public graph benchmarks demonstrate that our method significantly outperforms state-of-the-art approaches, achieving superior cross-task knowledge retention and multi-task continual learning performance.

Technology Category

Machine Learning: Graph-based Machine LearningData Mining & Knowledge Management: Graph Mining, Social Network Analysis & CommunityReasoning under Uncertainty: Graphical Models

Application Category

Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphsSemantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologiesSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

This paper addresses the challenge of incremental learning in growing graphs with increasingly complex tasks. The goal is to continuously train a graph model to handle new tasks while retaining proficiency in previous tasks via memory replay. Existing methods usually overlook the importance of memory diversity, limiting in selecting high-quality memory from previous tasks and remembering broad previous knowledge within the scarce memory on graphs. To address that, we introduce a novel holistic Diversified Memory Selection and Generation (DMSG) framework for incremental learning in graphs, which first introduces a buffer selection strategy that considers both intra-class and inter-class diversities, employing an efficient greedy algorithm for sampling representative training nodes from graphs into memory buffers after learning each new task. Then, to adequately rememorize the knowledge preserved in the memory buffer when learning new tasks, a diversified memory generation replay method is introduced. This method utilizes a variational layer to generate the distribution of buffer node embeddings and sample synthesized ones for replaying. Furthermore, an adversarial variational embedding learning method and a reconstruction-based decoder are proposed to maintain the integrity and consolidate the generalization of the synthesized node embeddings, respectively. Extensive experimental results on publicly accessible datasets demonstrate the superiority of method{} over state-of-the-art methods.

Problem

• Incremental learning in growing graphs with complex tasks
• Retaining proficiency in previous tasks via memory replay
• Enhancing memory diversity for better knowledge retention

Innovation

• Diversified Memory Selection and Generation framework
• Greedy algorithm for representative node sampling
• Adversarial variational embedding learning method