arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1903.07902cs.LGstat.ML
收藏

无监督网络表示学习的比较研究

A Comparative Study for Unsupervised Network Representation Learning

L3S Research center, Leibniz Universität(莱布尼茨大学L3S研究中心);University of Stavanger(斯塔万格大学)

Megha Khosla, Vinay Setty, Avishek Anand

摘要:

最近,图上的无监督网络表示学习(UNRL)方法取得了可观的进展,具有灵活的随机行走方法,新的优化目标和深度架构。然而,系统地比较嵌入以了解它们在不同图和任务中的行为并没有共同点。在本文中,我们从理论上组不同的方法下一个统一的框架和实证研究不同的网络表示方法的有效性。特别是,我们认为,大多数的UNRL方法显式或隐式模型和利用上下文信息的节点。因此,我们提出了一个框架,将各种方法-基于随机游走,矩阵分解和基于深度学习-转换为统一的基于上下文的优化函数。我们根据它们的相似性和差异对这些方法进行了系统的分组。我们详细研究了这些方法之间的差异,稍后我们将使用这些方法来解释它们的性能差异(下游任务)。我们进行了一项大规模的实证研究,考虑9个流行的和最近的UNRL技术和11个现实世界的数据集,具有不同的结构属性和两个常见的任务-节点分类和链接预测。我们发现,没有一种方法是一个明确的赢家,一个合适的方法的选择是由嵌入方法,任务和结构特性的基础图形的某些属性。此外,我们还报告了UNRL方法评价中的常见陷阱,并提出了实验设计和结果解释的建议。

英文摘要:

There has been appreciable progress in unsupervised network representation learning (UNRL) approaches over graphs recently with flexible random-walk approaches, new optimization objectives and deep architectures. However, there is no common ground for systematic comparison of embeddings to understand their behavior for different graphs and tasks. In this paper we theoretically group different approaches under a unifying framework and empirically investigate the effectiveness of different network representation methods. In particular, we argue that most of the UNRL approaches either explicitly or implicit model and exploit context information of a node. Consequently, we propose a framework that casts a variety of approaches -- random walk based, matrix factorization and deep learning based -- into a unified context-based optimization function. We systematically group the methods based on their similarities and differences. We study the differences among these methods in detail which we later use to explain their performance differences (on downstream tasks). We conduct a large-scale empirical study considering 9 popular and recent UNRL techniques and 11 real-world datasets with varying structural properties and two common tasks -- node classification and link prediction. We find that there is no single method that is a clear winner and that the choice of a suitable method is dictated by certain properties of the embedding methods, task and structural properties of the underlying graph. In addition we also report the common pitfalls in evaluation of UNRL methods and come up with suggestions for experimental design and interpretation of results.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑