Research On XML Keyword Search Results Sorting Problem

Because the structured query method can return the accurate search results to users,users often use structured query language XPath and XQuery when query XML data.But the structured query method not only asks users to learn the complex query language and write accurate query expressions,but also requires users to know about the underlying structure of XML documents.As a simple,practical and efficient way to get information,keyword search technology for XML data come into being.Compared with the traditional XML structured query method,XML keyword query not only provides users with a simple and practical query interface that similar to the search engine,and the query process is very simple and flexible.XML keyword query is a friendly and convenient way to search information.In XML keyword query,users can get their own target search results only need to enter query keywords related to the query.Simple query keywords in XML keyword query are not enough to express the user's real search intention.Facing the ambiguity of keywords semantic information,this paper set the weight value for keywords to reflect the user's real search intention.In this paper,we propose a new result ranking method combine the relationship between nodes with the weight value in the weight relationship tree.The weight value of the query keywords is determined firstly in this result ranking method.Then the weight value of keyword nodes in the weight relationship tree is set according to the weight value of query keywords,and set the weight value of other non-keyword nodes according to the relationship between nodes.Finally,the ranking algorithm calculates the importance of each SLCA depends on the weight value of the nodes in weight relationship tree and the structural relationship between nodes,the search results of the XML keyword query sorted based on the importance of SLCA.In addition to results ranking method based on weight value of XML query keywords,the sorting problem of XML keyword search results should take into account users' different preferences for different content of the query.System can guess the user's query preferences by collecting the history records of searching,then analyze the query history records of users and calculate the freshness and interest of result fragments,and calculate the importance of each query result based on the freshness and interest of the resulting fragments after normalizing,then sort the results according to the importance of searching results.Finally,orderly and preferential search results returned to the users.In order to verify the validity of the ranking algorithm proposed in this paper,the XML keyword query system is designed and implemented.In this system,the results returned by the XML keyword query are tested with two different ranking algorithms.Two kinds of ranking algorithms about XML keyword query are experimented on two different data sets,and the experimental results are compared and analyzed.This experiment verifies the validity of the ranking algorithm from the aspects of rank accuracy and sequencing efficiency.