MOHONE: Modeling Higher Order Network Effects in KnowledgeGraphs via Network Infused Embeddings
This work addresses the problem of limited connectivity modeling in knowledge graphs for researchers and practitioners, offering a generic framework that is incremental but provides strong specific gains.
The paper tackles the limitation of local views in knowledge graph embedding methods by introducing MOHONE, a framework that models higher-order network effects to capture varying degrees of connectivity, resulting in consistent and significant improvements in link prediction performance, such as at least 4 points or 17% gains for methods like TRANSE and DISTMULT.
Many knowledge graph embedding methods operate on triples and are therefore implicitly limited by a very local view of the entire knowledge graph. We present a new framework MOHONE to effectively model higher order network effects in knowledge-graphs, thus enabling one to capture varying degrees of network connectivity (from the local to the global). Our framework is generic, explicitly models the network scale, and captures two different aspects of similarity in networks: (a) shared local neighborhood and (b) structural role-based similarity. First, we introduce methods that learn network representations of entities in the knowledge graph capturing these varied aspects of similarity. We then propose a fast, efficient method to incorporate the information captured by these network representations into existing knowledge graph embeddings. We show that our method consistently and significantly improves the performance on link prediction of several different knowledge-graph embedding methods including TRANSE, TRANSD, DISTMULT, and COMPLEX(by at least 4 points or 17% in some cases).