图分类研究综述

图数据广泛存在于现实世界中,可以自然地表示复合对象及其元素之间的复杂关联.对图数据的分类是一个非常重要且极具挑战的问题,在生物/化学信息学等领域有许多关键应用,如分子属性判断,新药发现等.但目前尚缺乏对于图分类研究的完整综述.首先给出图分类问题的定义和该领域的挑战;然后梳理分析了两类图分类方法—基于图相似度计算的图分类方法和基于图神经网络的图分类方法;接着给出了图分类方法的评价指标、常用数据集和实验结果对比;最后介绍了图分类常见的实际应用场景,展望了图分类领域的未来研究方向并对全文进行总结.

Survey on Graph Classification

Graph data, as a kind of widely-existed data in the real world, naturally represent complex interactions between elements of composite objects. The classification of graph data is a very important and extremely challenging reasearch problem. There are many key applications in the fields of bio/chemical informatics, such as molecular attribute classification and drug discovery. However, there still lacks a comprehensive review of research on graph classification. In this survey, we first formulate the problem of graph classification and describe the main challenges of this problem; We then categorize graph classification method into similarity-based methods and graph neural network based methods. Moreover, evaluation metrics for graph classification, benchmark datasets and comparison results are given; Finally, the application scenarios of graph classifications are summarized and we also discuss the research trends of graph classification.