Answering Durable Skyline Queries on Multidimensional Time Series Data Using Grid-Based Approach

Data shows an essential role in everyday life, particularly in the field of data analysis. It serves as a vital component for obtaining information used in decision-making processes. The rapid advancement of science and technology in data analysis has influenced the methods of data retrieval to be more efficient.One of the data engineering methods is the Skyline Query. The Skyline Query is a search method that generates interesting objects, referred to as skyline objects, which are superior to other objects and not worse than any other object in all attributes. However, the Skyline Query has a limitation as it does not consider time variables in its calculation algorithm, thus making it unable to process time interval-based queries.This article aims to address the issue of time interval-based Skyline queries on historical multidimensional data with a time series by modeling the Durable Skyline Query and designing a framework algorithm to process such queries. Two types of algorithms are created and compared: the grid-based and the naïve one. The effectiveness and efficiency of the algorithms are tested using real and synthetic data.Based on the experimental results, the grid algorithm outperforms the naive algorithm in independent (IND) and Forest Cover Type (FC) data, as it provides query results with computation time two to five times faster than the naïve algorithm. Several parameters that have an impact on precomputing and query processing time were also thoroughly observed in this work.