FastGCL: Fast Self-Supervised Graph Learning for Service Understanding

Huizhe Zhang,Wangbin Sun,Liang Chen

DOI: https://doi.org/10.1109/icws62655.2024.00042

2024-01-01

Abstract:Benefit from the widespread implement of data collection technologies, a large number of data services are emerged in various industries like finance and energy. The complex internal connectivity within these services can be naturally modeled as the graph-structured data, which facilitates services’ understanding based on graph mining. However, the heavy label reliance limits the development of traditional graph learning methods on data services. Graph contrastive learning (GCL), as a popular graph representation learning approach without relying on manual labels, has recently achieved a non-negligible effect. Thus, for alleviating above problems, we propose a simple yet effective method named FastGCL containing a better contrastive scheme which can be tailored to the characteristics of graph neural networks (e.g., neighborhood aggregation). Specifically, by constructing weighted-aggregated and non-aggregated neighborhood information as positive and negative samples respectively, FastGCL identifies the potential semantic information of data without disturbing the graph topology and node attributes, resulting in faster training and convergence speeds. Extensive experiments have been conducted on node classification and graph classification tasks, showing that FastGCL has competitive classification performance and significant training speedup compared to existing state-of-the-art methods.