Design And Implementation Of Interactive Visualization System Of Distributed Large Graph

With the rapid development of the Internet,various Internet applications have brought about tremendous changes in People's Daily necessities and in the politics,economy and society of various countries.Behind this environment,huge amounts of data of various types are generated at an alarming rate.People have entered the era of big data.The huge energy of data in this era can be said to be amazing,from accurate recommendation,blind date matching to artificial intelligence and so on,and big data is everywhere.Various data mining and analysis technologies play an important role in the popularity of big data.Among these various types of data,graph data occupies a place.Graph provide a simple,powerful,and effective way to model data.In the era of big data,data that can be modeled with graphs can be seen everywhere.Specifically,it mainly comes from various networks.Examples include telecommunications networks,computer networks,biological networks,cognitive and semantic networks,transportation networks,crime networks,citation networks,and social networks.Like mining and analyzing other types of data,there are many mature techniques for analyzing graph data.It falls into two categories.One is the aggregation technology of graph.According to the requirements,the analyst aggregates the vertices or edges in the graph according to certain rules to achieve the analysis purpose,such as some community discovery technologies.The second is the visualization technology of the graph,which involves multiple parts such as layout algorithm,visual chaos reduction and interaction,etc.,through which the nodes and edges of the graph are displayed on the screen and interactive analysis function is provided.The visualization of large graphs is one of the hot spots and difficult points in the field of graph visualization.In the research and design of this technology,this paper mainly combines the current popular framework and the technology proposed by my postgraduate tutor Dr.HuiJu Wang.Based on this,this paper mainly does the following work:One is to summarize and sort out the research status at home and abroad.At home and abroad,graph analysis and research are mainly carried out based on graph data or graph structure.Graph aggregation and other technologies are developed on data pattern,while visualization is carried out based on figure of data.The second is to study and research in detail and sort out the three frameworks or techniques adopted in this paper.The first technology is based on the interactive exploration technology of pile nodes proposed by Dr.HuiJu Wang.Combining the graph aggregation technology and the graph visualization method,this theory divides the vertices in the graph into pile nodes and non-pile nodes.The pile nodes are displayed,while the non-pile nodes are not directly displayed by subgraph extraction and aggregation,so as to make the visualization simple and effective.The second framework is Spark based graphics computing framework GraphX.GraphX combines the advantages of data parallelism and graph parallelism,and effectively expresses graph computation based on Spark.The third framework is GraphStream,a Java library for graph processing,which focuses on the dynamic processing of graphs and is very suitable for the visual presentation and interaction analysis of dynamic graphs.Third,on the basis of detailed investigation and analysis,an interactive visual exploration system of graph is designed and implemented.