Efficient Skyline Maintenance for Streaming Data with Partially-Ordered Domains

Abstract

Abstract. We address the problem of skyline query processing for a count-based window of continuous streaming data that involves both totally-and partially-ordered attribute domains. In this problem, a fixedsize buffer of the N most recent tuples is dynamically maintained and the key challenge is how to efficiently maintain the skyline of the sliding window of N tuples as new tuples arrive and old tuples expire. We identify the limitations of the state-of-the-art approach STARS, and propose two new approaches, STARS + and SkyGrid, to address its drawbacks. STARS+ is an enhancement of STARS with three new optimization techniques, while SkyGrid is a simplification STARS that eliminates a key data structure used in STARS. While both new approaches outperform STARS significantly, the surprising result is that the best approach turns out to be the simplest approach, SkyGrid.

Efficient Computation of Skyline Queries Over a Dynamic and Incomplete Database

“…This section presents the experimental results of the DyIn-Skyline solution in processing skyline queries over a dynamic and incomplete database, in which the changing state of the database is due to a data manipulation operation(s) (insert, delete or update a data item(s) Effect of Changing Rate -One of the factors that has significant effect on the performance of skyline algorithms in processing skyline queries over a dynamic database is the changing rate of the database. In this section, we illustrate the experimental results of our proposed solution and the previous algorithms for both the synthetic and real data sets with respect to the number of pairwise comparisons and processing time, by varying the changing rate from 5% -30% as applied in the previous studies [26], [35] with 20% incompleteness rate. The number of dimensions is fixed to 15, 13, 4, and 16 for the synthetic, NBA, MovieLens, and stock market data set, respectively.…”

“…These preference queries employ preference evaluation techniques, have achieved significant success, as they are widely used in applications related to multi-criteria decision support. During the two past decades, several preference evaluation techniques have been introduced, among them are: top-k [30], skyline [8], [10], [11], [12], [13], [20], [25], [29], [33], [35], k-dominance [5], top-k dominating [19], and kfrequency [6].…”

Efficient Skyline Computation Over an Incomplete Database With Changing States and Structures

“…With the rapid growth of decision support systems and the increasing size of multidimensional data have witnessed an abundance of skyline algorithms being proposed for data processing in order to retrieve useful insights. These variants of skyline algorithms are introduced to deal with different characteristics of data, such as uncertain data [23], [27], [29], [39], [40], [41], [51], incomplete data [3], [6], [13], [14], [18], [19], [20], [24], [28], [33], [49], [50], encrypted data [8], [31], and streaming data [1], [15]; while others are based on the platform being considered like distributed database [2], cloud computing [22], [32], road networks [16], and others.…”