Construction of 5G Knowledge Graph Based on Multidimensional Relationship Extraction

As 5G technology surges forward, the need for network experts to grasp communication protocols escalates. Building a 5G protocol knowledge graph is now crucial. This study introduces a model that unifies entity and relationship encoding for triple extraction, enhancing complex relationship processing in 3GPP protocols. In addition, traditional relationship extraction methods often overlook the qualifying attributes of relationship triples, such as time, quantity, which are crucial for capturing the complexity of knowledge graphs. It also addresses the oversight of relationship triple attributes by proposing a multidimensional relationship extraction model. The method boosts triple extraction precision and supports extracting richer knowledge graph structures through quintuple extraction from texts.