Study On Skyline Aggregation Queries

The Skyline has been proposed as an important operator for the field of Special Database related on multi-criteria decision making, data mining, test and observation, user-fancy query and visualization, which is a common and complex query. Nowadays, the researches on it refer to the simple skyline query only, which can not satisfy the actual application needs. Based on it, this paper brings about the skyline aggregation query that can satisfy the complex application needs, and proposes corresponding query processing ways.Based on the common situation this paper proposes the basic algorithm, which called aggregation first algorithm AACN (Aggregate All, Compute Next) as we analyze the different aggregation function. The basic idea of AACN algorithm is that Firstly aggregate all the dataset based on time, secondly do the skyline query on them. Based on it we analyze the algorithm and bring forwards another two algorithms, which called time first algorithm CETAN (Compute on Every Time, Aggregate Next) and algorithm based on filter strategy ABT (Aggregation Based on Time). Among them CETAN algorithm is that firstly do the skyline query on every time, secondly aggregate the result list; while the key idea of ABT algorithm is that using filter to filtrate and update the dataset constantly, get our results in the end, and this algorithm allows the dynamic updating on dataset.At last using three types of datasets, which called Correlated data, anti-Correlated data and independent data respect, experiment simulates and tests the three algorithms. And the Experimental results validate that our approaches are feasible and improve our query on precision and variety.