Improved Top-k Query Processing on Uncertain Data

Abstract

Bottom-up algorithm, which is one of the two probabilistic Top-k query algorithms, was improved. The core of the bottomup algorithm is the iteration on the three courses of bounding, pruning,and refining towards the objects and instances. The main contribution is to change the iteration on instances of objects one by one into iterating all the instances of objects from the superior to the inferior;and to transform the condition and sequence of pruning in order to make the pruning more effective. Theoretical analysis and experimental results show that the algorithm efficiency could be obviously increased by about 20%.