CSQUiD: an index and non-probability framework for constrained skyline query processing over uncertain data

Area of Science:

• Database Management
• Data Mining
• Information Retrieval

Background:

• Data uncertainty is prevalent in modern databases, posing challenges for efficient skyline query processing.
• Existing methods, such as probability-based and median-based approaches, are unsuitable for high-dimensional uncertain data.
• These methods suffer from intensive computations or impracticality in real-world applications.

Purpose of the Study:

• To introduce a novel, non-probability framework named Constrained Skyline Query processing on Uncertain Data (CSQUiD).
• To reduce computational time for processing constrained skyline queries over uncertain high-dimensional data.
• To provide an efficient alternative to existing methods for handling data uncertainty in databases.

Main Methods:

• CSQUiD utilizes an X-tree indexing structure to construct Minimum Bounding Rectangles (MBRs) for uncertain data objects.
• It analyzes only objects within dominant MBRs, avoiding exhaustive scanning of the entire dataset.
• The Fuzzification approach is employed to identify exact values for continuous range data within dominant MBRs.

Main Results:

• Extensive experiments on real and synthetic datasets demonstrate CSQUiD's effectiveness.
• CSQUiD significantly outperformed the CIS algorithm and SkyQUD-T framework in terms of pairwise comparisons (44.07% and 57.15% improvement).
• CSQUiD achieved substantial improvements in CPU processing time over CIS (27.17%) and SkyQUD-T (18.62%).

Conclusions:

• CSQUiD offers a computationally efficient and practical solution for constrained skyline queries on uncertain high-dimensional data.
• The framework's performance advantages make it a viable alternative for modern database applications.
• CSQUiD effectively addresses the limitations of previous approaches in handling data uncertainty.