Communicative Message Passing for Inductive Relation Reasoning
This work is significant for researchers and practitioners working with knowledge graphs, particularly for improving the inductive generalization capabilities of relation prediction models, which is a common bottleneck in real-world applications. It offers an incremental improvement over existing subgraph-based methods.
This paper addresses inductive relation prediction in knowledge graphs, where models need to predict missing relationships for unseen entities. The authors propose CoMPILE, a Communicative Message Passing neural network, which improves upon existing subgraph-based methods by better handling directed subgraphs and strengthening relation information flow. CoMPILE achieves substantial performance gains over state-of-the-art methods on benchmark datasets in inductive settings.
Relation prediction for knowledge graphs aims at predicting missing relationships between entities. Despite the importance of inductive relation prediction, most previous works are limited to a transductive setting and cannot process previously unseen entities. The recent proposed subgraph-based relation reasoning models provided alternatives to predict links from the subgraph structure surrounding a candidate triplet inductively. However, we observe that these methods often neglect the directed nature of the extracted subgraph and weaken the role of relation information in the subgraph modeling. As a result, they fail to effectively handle the asymmetric/anti-symmetric triplets and produce insufficient embeddings for the target triplets. To this end, we introduce a \textbf{C}\textbf{o}mmunicative \textbf{M}essage \textbf{P}assing neural network for \textbf{I}nductive re\textbf{L}ation r\textbf{E}asoning, \textbf{CoMPILE}, that reasons over local directed subgraph structures and has a vigorous inductive bias to process entity-independent semantic relations. In contrast to existing models, CoMPILE strengthens the message interactions between edges and entitles through a communicative kernel and enables a sufficient flow of relation information. Moreover, we demonstrate that CoMPILE can naturally handle asymmetric/anti-symmetric relations without the need for explosively increasing the number of model parameters by extracting the directed enclosing subgraphs. Extensive experiments show substantial performance gains in comparison to state-of-the-art methods on commonly used benchmark datasets with variant inductive settings.