Research On The Construction And Application Of Scientific Knowledge Graph

With the fast development of computer science,many research fields have been proposed and each research field has many research topics.Excavating important research topics and analyzing those topics' variation tendency is of great importance.There have been some papers for this issue but they only use quantity of papers to value the importance of one research topic which does not have enough reliability.In this paper,we will use not only the quantity of papers and the importance to measure one research topic.Visualization tool CiteSpace is used to visualize and analyze the papers collected in the field of services computing from the paper indexing database Web of Science(WOS)from 2003-2017.Using Knowledge graphs,the study shows the keywords co-occurrence network and reveals the hot research topics in this field.And based on social network analysis methods,the study will use betweenness centrality to measure the importance and influence of the nodes in the network.The keywords that have high frequency and high betweenness centrality will be regarded as the most important nodes in the network and they are the most important research areas in this field.In this resarch,CiteSpace will be used to analyze term frequency and generate keywords co-occurrence network and a program has been designed and developed to compute node centrality in this research.Based on the experiment,it is revealed that web services,cloud computing,big data,ontology and semantic web are the most important research topics in recent ten years.In addition,there are similarities between the research situation at home and abroad.