Network security knowledge graphs constructed from cybersecurity knowledge often suffer from incompleteness and insufficient utilization of available information. To address these issues, this study proposes SynerKGC, a knowledge graph completion model that integrates semantic, structural, and type information to improve the performance of cybersecurity knowledge graph completion tasks. Specifically, TransE is employed to model the structural features of triples, while a graph convolutional network (GCN) is used to capture topological features by incorporating entity type information. A gating mechanism is then applied to fuse these structural features, thereby enhancing type-aware structural representations. In addition, SecBERT is used to extract semantic vectors from entity and relation descriptions as semantic representations, which are further organized into triple sequence representations and fed into SecBERT with Adapters to capture contextual semantic features. The extracted semantic features are subsequently concatenated and fused with structural representations. Furthermore, InfoNCE is introduced to align semantic and structural representations, strengthening the overall representation of entities. Experimental results on both a self-constructed cybersecurity dataset and the general benchmark dataset WN18RR demonstrate that SynerKGC achieves superior performance on the link prediction task, outperforming most existing baseline models.