Google Knowledge Graph Hotspot Frontier Analysis And Research On Improvement Of Co-word Network Clustering

Field hotspots and frontier research provide researchers with a clear focus on research and future trends by analyzing and visualizing the processing of keywords in the literature.As a core technology that makes machines have cognitive ability,Google Knowledge Graph has been developed for many years,but there has never been related research work in this field.Therefore,it is necessary to conduct research on hotspots and frontier of Google Knowledge Graph.However,due to the confusion between the two concepts of "scientific knowledge map" in the field of library and information and "Google Knowledge Graph",it has certain difficulties in obtaining and researching relevant data.Therefore,it is necessary to manually screen the data of Google Knowledge Graph.In addition,because there are many problems in the analysis process of the relevant analytical tools on the market,such as the poor effect of network clustering,lack of evaluation of effects,etc.,it is necessary to design more appropriate related algorithms for these data and apply them to the field hotspot research and trend analysis of Google Knowledge Graph.Finally,through the calculation of the burst words of Google Knowledge Graph,and then visualize them to the entire time dimension,the frontiers of Google Knowledge Graph are derived.This dissertation proposes an improved spectral clustering algorithm based on density sensitivity,which solves the problem of poor clustering effect of Co-word network and solves the problem of lack of clustering effect evaluation in existing analysis tools.In the comparative experiment,its normalized mutual information increased by 0.131,and the clustering error rate index decreased by 0.145.Then the density-sensitive improved spectral clustering algorithm is applied to the co-word network clustering algorithm to analyze the relevant hotspots of Google Knowledge Graph,and compare the analysis with the existing and popular Citespace tools to analyze the results of Google Knowledge Graph.Finally,this dissertation innovatively applies the burst detection algorithm to the trend analysis of Google Knowledge Graph,then realizes the prediction and verify the future research direction of Google Knowledge Graph.