Research On IoT Semantic Modeling Based On Graph Representation Learning

With the quick development of the Internet of Things(IoT),the number of connected things and their interactions continue to increase.The advent of emerging communication technologies such as Internet Protocol Version 6(IPv6)and fifthgeneration(5G),provides sufficient network addresses and improves the efficiency of data transmission.These technologies further accelerate the growth of connected things.The interaction and communication among a large number of things generate an enormous amount of context-aware data that is semantically diverse.Transforming these data into knowledge and represent it as the machine-understandable form will help improve the scalability and interoperability of the IoT system,thereby promoting the development of IoT applications and the processing and fusion of IoT data.However,traditional data representation approaches such as semantic annotation,ontology,and semantic web technology are rule-based,which lack flexibility and adaptability when applied to IoT.To address the challenge,this paper mainly focuses on the problem of semantic representation,which is essential for processing and fusion of IoT data.To serve as a bridge,we propose a high-level framework,namely Things2 Vec,which aims to produce the latent semantic representations from the interaction of things though the graph representation learning technique.These semantic representations benefit various IoT semantic analysis tasks such as the IoT service recommendation and automation of things.In Things2 Vec,we utilize the graph to model the function sequence relationships that are generated by the interaction of things,which is called the IoT context graph.Since these function sequence relationships are heterogeneous in terms of semantics,it causes general graph representation learning methods to fail to learn complete information.Thus,we propose a biased random walk procedure,which is tailored to capture the neighborhoods of nodes with different types of semantic relationships.To demonstrate the effectiveness of the proposed framework,we conduct a multi-label classification experiment on the real-world IoT dataset IFTTT.Experimental results show that three general graph representation learning methods all show good performance based on our proposed framework.Furthermore,Things2 Vec with just 20% labeled nodes achieves 3%~13% improvements in terms of Micro-F1 and 3%~37% gains in terms of Macro-F1 over other compared methods.That proves that the proposed method can effectively capture the semantic relationship between contextaware data in IoT.