Joint Graph Embedding and Alignment with Spectral Pivot
Joint Graph Embedding and Alignment with Spectral Pivot
复制标题
DOI：
10.1145/3447548.3467377
复制
发表时间：
2021-08
期刊：
影响因子：
--
通讯作者：
Paris A. Karakasis;Aritra Konar;N. Sidiropoulos
中科院分区：
文献类型：
--
作者：
Paris A. Karakasis;Aritra Konar;N. Sidiropoulos
图形是强大的抽象，自然地捕捉我们相互关联的世界中的关系财富。针对图挖掘中的核心问题--图对齐问题，提出了一种新的方法.经典（例如，谱）方法对两个图使用固定嵌入来执行对齐。相比之下，所提出的方法固定的“目标”图的嵌入和联合优化的嵌入变换和对齐的“查询”图。提出了一种交替优化算法，用于计算高质量的近似解，并与使用基准真实世界图的当前最先进的图对齐框架进行比较。结果表明，所提出的配方可以提供显着的增益在匹配精度和鲁棒性噪声相对于现有的解决方案，这个困难，但重要的问题。
Graphs are powerful abstractions that naturally capture the wealth of relationships in our interconnected world. This paper proposes a new approach for graph alignment, a core problem in graph mining. Classical (e.g., spectral) methods use fixed embeddings for both graphs to perform the alignment. In contrast, the proposed approach fixes the embedding of the 'target' graph and jointly optimizes the embedding transformation and the alignment of the 'query' graph. An alternating optimization algorithm is proposed for computing high-quality approximate solutions and compared against the prevailing state-of-the-art graph aligning frameworks using benchmark real-world graphs. The results indicate that the proposed formulation can offer significant gains in terms of matching accuracy and robustness to noise relative to existing solutions for this hard but important problem.