Font Size: a A A

Research And Implementation Of Keyword Query Algorithm For RDF Graph

Posted on:2019-01-17Degree:MasterType:Thesis
Country:ChinaCandidate:Y DingFull Text:PDF
GTID:2428330542994441Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
At present,RDF data sets are widely used in various fields,and user's query requirements for RDF data are also increasing.However,the keyword search methods which have been put forward based on RDF graphs can't realize to query the key words on the edge,and RDF graphs can be converted to the RDF bipartite graphs in order to achieve the keywords of edge query but increasing the number of vertices in the graphs which affects the efficiency of the query.At the same time,the problem of insufficient storage space and low query efficiency is found in the centralized environment.To solve above problems,this paper studies the existing methods about keyword query and graph segmentation and the keyword query method based on index is proposed,so as to realize the high efficiency of RDF data query.The main work of this paper includes:(1)In order to realize the query of vertices and edges of the RDF data graph at the same time,a query method of RDF graph based on double indexing mechanism is proposed.This method first converts RDF data into RDF graph,and uses the adjacency label matrix of RDF graph to construct a vertex index and edge index to improve query efficiency of keywords;Secondly,the keyword's location of the double index is determined by the string matching method,and the vertex and edge query are realized according to the relationship between the two indexes.Finally,the correlation evaluation function is used to sort the query result subgraph and output top-k query result.(2)In order to realize parallel query of keywords,an edge segmentation algorithm based on RDF sentence graph is proposed,which uses the multi-level segmentation method to segment the RDF data graph.Firstly,the RDF data is transformed into an RDF sentence graph,and an RDF sentence graph is segmented using the minimum degree vertex algorithm;Secondly,the edge segmentation based on RDF sentence graph is equivalent to the vertex segmentation of the RDF graph,thus the intersection of each subgraph is the vertex cut;Finally,using the mapreduce parallel computing framework to achieve parallel keywords query.The experiment shows that the performance of the query method of RDF graph based on the double indexing mechanism and the edge segmentation algorithm based on RDF sentence graph is better than the comparative methods.
Keywords/Search Tags:keyword query, RDF graph, double index, RDF sentence graph, multi-level segmentation, parallel query
PDF Full Text Request
Related items