Are Message Passing Neural Networks Really Helpful for Knowledge Graph Completion?
Are Message Passing Neural Networks Really Helpful for Knowledge Graph Completion?
复制标题
DOI：
10.18653/v1/2023.acl-long.597
复制
发表时间：
2022-05
期刊：
影响因子：
--
通讯作者：
Juanhui Li;Harry Shomer;Jiayu Ding;Yiqi Wang;Yao Ma;Neil Shah;Jiliang Tang;Dawei Yin
中科院分区：
文献类型：
--
作者：
Juanhui Li;Harry Shomer;Jiayu Ding;Yiqi Wang;Yao Ma;Neil Shah;Jiliang Tang;Dawei Yin
知识图（KGs）促进了各种各样的应用。尽管在创建和维护方面付出了巨大努力，但即使是最大的幼儿园也远未完工。因此，幼儿园完成（KGC）已成为幼儿园研究的最重要的任务之一。最近，这个领域的大量文献都集中在使用消息传递（图）神经网络（MPNN）来学习强大的嵌入。这些方法的成功自然归功于MPNN在更简单的多层感知器（MLP）模型上的使用，因为它们具有额外的消息传递（MP）组件。在这项工作中，我们发现令人惊讶的是，简单的MLP模型能够实现与MPNN相当的性能，这表明MP可能不像以前认为的那样重要。通过进一步的探索，我们发现仔细的评分函数和损失函数设计对KGC模型的性能有更大的影响。这表明在以前的工作中，评分函数设计，损失函数设计和MP的合并，与有前途的见解，关于国家的最先进的KGC方法的可扩展性今天，以及仔细注意更合适的MP设计KGC任务的明天。
Knowledge graphs (KGs) facilitate a wide variety of applications. Despite great efforts in creation and maintenance, even the largest KGs are far from complete. Hence, KG completion (KGC) has become one of the most crucial tasks for KG research. Recently, considerable literature in this space has centered around the use of Message Passing (Graph) Neural Networks (MPNNs), to learn powerful embeddings. The success of these methods is naturally attributed to the use of MPNNs over simpler multi-layer perceptron (MLP) models, given their additional message passing (MP) component. In this work, we find that surprisingly, simple MLP models are able to achieve comparable performance to MPNNs, suggesting that MP may not be as crucial as previously believed. With further exploration, we show careful scoring function and loss function design has a much stronger influence on KGC model performance. This suggests a conflation of scoring function design, loss function design, and MP in prior work, with promising insights regarding the scalability of state-of-the-art KGC methods today, as well as careful attention to more suitable MP designs for KGC tasks tomorrow.