Incremental Parallel Graph Query For Large Graph Data

Along with the highly advanced information technology and the Internet,a large amount of data has been generated and accumulated in various industries,especially the computer industry,and massive data continues to be generated and accumulated.Because of the size of the data and their complexity,massive amounts of data are difficult to process on a single machine,traditional data storage and management tools and processing algorithms and data processing have become unusable or the efficiency has become extremely low,unacceptable.There is an urgent need to query,extract and analyze the information people need from massive data in order to make further decisions quickly or to develop a work plan for the next step.Graphs are widely used to describe complex data structure models,for example,in many applications such as social networks and the Semantic Web.So many queries and searches can be constructed into a graph,so the value data extraction and query problems of massive data are transformed into graph query problems.However,in the real world,not only the structure of the graph has been changing,but how to obtain valuable information in the changed large graph data has become an urgent problem.In this paper,an incremental parallel graph query algorithm is proposed based on the research of dynamic large graph data graph pattern query.The shortest distance in dynamic graph is maintained effectively and the traditional graph pattern matching algorithm is implemented,which is parallel by using MapReduce programming mode of Hadoop distributed computer platform.The shortest distance of all pairs of vertices dynamic large graph data is maintained incrementally in parallel,and parallel graph pattern query is realized.Firstly,this paper introduces the research status of reachable index and query of graph pattern at home and abroad,and also introduces the effective maintenance algorithm of the shortest distance in dynamic graph and the distributed computing platform Hadoop and its components.Then,the incremental parallel graph query algorithm is designed and implemented.Finally,the effectiveness and performance of the algorithm are analyzed through a number of practical data sets.The experimental results show that the incremental parallel graph query algorithm provides an effective way to solve the query problem of dynamic large graph data pattern.