Research On Multidimensional Data Query Processing Based On User Preferences

In the era of big data,data query tends to be personalized,and query users have higher requirements for the quality of query results set.It is very important to select the most interested objects from multidimensional data sets based on user preferences.In the new era,data is massive and multidimensional.The existing multidimensional data query has defects in supporting personalized query and ensuring the quality of the result set.Therefore,it is of great significance to research multidimensional data query under new query preferences.The existing multidimensional data query is divided into two categories based on query objects: single object query and group object query.This paper focuses on related issues of interactive object query and role combination under group selection under dynamic preferences.(1)The analysis of one of the most recent branches of multidimensional data query is available,and there is an association between iterative queries on the same problem.But the difference between the query user preferences may change dynamically,and the existing interactive query algorithm between the default query preference is fixed,I propose a support dynamic preference query processing algorithm IMQD;quality metrics to define a set of results,the value and value is more excellent in the range of [0,1].The IMQD algorithm obtains the preference threshold by interacting with the user,and then updates the result set to make the result set higher,in which the preference threshold supports the dynamic adjustment of the user.The correctness and effectiveness of IMQD algorithm is verified on simulated and real datasets.Once user preferences change,the algorithm can adjust the result set in milliseconds.(2)Based on the current situation of existing group object query algorithm,which results in uncontrollable size of the result set based on Skyline operator,and connecting with the query scenes of a large number of optimal group objects in real life,users usually have the need to acquire and obtain only one optimal group object.Combined with the optimal single object Top-k query,we propose the role group optimization problem under quantitative preferences,and propose the corresponding query processing algorithm--GQBRs algorithm.The algorithm generates the final query results in three steps.First,the candidate set is determined based on the group member preferences,and then the candidate group objects are generated based on the filtered data set.Finally,GQBRs sorts all the candidate group objects and returns the set of object groups of the specified size.It is proved that the result set of GQBRs algorithm is controllable on the simulation and real data sets.(3)The existing multidimensional data query processing algorithm,I design an query processing framework for all existing multidimensional data query,and verifie in the IMQD algorithm and GQBRs algorithm,and base on this framework according to the data set,query preferences,query processing,display the result set of four sub modules is realized the prototype system.The experiment validates the correctness of the extension of the IMQD and GQBRs algorithms in the prototype system.