Research On Query Processing Algorithms In Uncertain Data Based DRA

With the large number of uncertain data produced, how to get the result of the Top-kquery in uncertain database is becoming an urgent problem. Because the special dimension ofprobability exists in uncertain data, making it is very different with the certainty traditionaldata in processing methods. It has a great significance to solve the query problem in uncertaindata accurately and efficiently.First, the causes of how to generate uncertainty data, and the characteristics of uncertaindata, compare it with deterministic data and find the differences of them. Then summarize theexisting query processing models and methods for the uncertain data, and point out theproblems in these models explain the corresponding methods are how to solve these problems,as well as advantages and disadvantages of existing methods. This paper not only study theexisting Top-k methods of processing uncertain data, but also study the relationship betweenSkyline queries and queries and Top-k queries. use the existing methods of processing Skylinequeries to solve the Top-k queries. Finally, we proposed use the DRA (dominate relationshipanalysis) of the tuples to improve existing methods, the idea determine the controlrelationships of the tuples by compare the score and probability, making some of thecontrolled tuples excluded directly before query processing, do not appear in Top-k querycalculation. Witch can make the existing query methods more efficiently and accurately.This paper presents the idea of using the DRA (dominate relationship analysis) of thetuples to delete those tuples can not be the result, making the query processing be moresimple.On the other hand, for the frequently updated database, this method showed moreexcellent effect. It could determine whether the change tuple will make an impact on the queryresults, thus saving the time and space for re-query. Experimental results show that the methodwe proposed can be better meet the user’s requirements for fast and accurate get the queryresult, and the search efficiency is also getting improved.