Multi-Scale Dynamic Convolutional Network for Knowledge Graph Embedding

• Published in: IEEE transactions on knowledge and data engineering Vol. 34; no. 5; pp. 2335 - 2347
• Main Authors: Zhang, Zhaoli, Li, Zhifei, Liu, Hai, Xiong, Neal N.
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.05.2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: complex relations; Computational modeling; Computer architecture; Convolution; convolutional network; Embedding; Feature extraction; Graphs; Knowledge bases (artificial intelligence); Knowledge engineering; knowledge graph embedding; Knowledge graphs; Knowledge representation; link prediction; Predictive models; Semantics
• ISSN: 1041-4347; 1558-2191
• DOI: 10.1109/TKDE.2020.3005952

Knowledge graphs are large graph-structured knowledge bases with incomplete or partial information. Numerous studies have focused on knowledge graph embedding to identify the embedded representation of entities and relations, thereby predicting missing relations between entities. Previous embedding models primarily regard (subject entity, relation, and object entity) triplet as translational distance or semantic matching in vector space. However, these models only learn a few expressive features and hard to handle complex relations, i.e., 1-to-N, N-to-1, and N-to-N, in knowledge graphs. To overcome these issues, we introduce a multi-scale dynamic convolutional network (M-DCN) model for knowledge graph embedding. This model features topnotch performance and an ability to generate richer and more expressive feature embeddings than its counterparts. The subject entity and relation embeddings in M-DCN are composed in an alternating pattern in the input layer, which helps extract additional feature interactions and increase the expressiveness. Multi-scale filters are generated in the convolution layer to learn different characteristics among input embeddings. Specifically, the weights of these filters are dynamically related to each relation to model complex relations. The performance of M-DCN on the five benchmark datasets is tested via experiments. Results show that the model can effectively handle complex relations and achieve state-of-the-art link prediction results on most evaluation metrics.