Horizontal Aggregation in Relational Database Management Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional database systems produce unwieldy vertical outputs for group-by operations, making it difficult to understand and analyze large datasets in applications like OLAP and data mining, requiring significant effort to summarize detailed aggregation outputs.
Innovation Solution
Implementing horizontal aggregation, where multiple aggregate values are arranged in a two-dimensional matrix format, allowing outputs of grouping and aggregate operations to be more easily understood and analyzed, using the HAGG() function to compute aggregates based on specified grouping and sub-grouping columns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional vertical aggregation output format is used, then complete aggregation results are produced, but the output becomes difficult to understand and analyze when dealing with large datasets
Solution Approach 1:
The patent transforms the conventional vertical aggregation output (one aggregate value per row) into a horizontal aggregation output format where multiple aggregate values are arranged in a two-dimensional matrix structure. This dimensionality change allows the same aggregation results to be presented in a format that is more suitable for OLAP and data mining applications, improving both understandability and analytical efficiency without losing any information.
2Quantity of substance
If detailed aggregation outputs are produced for large datasets, then comprehensive analysis data is available, but significant time and effort are required to summarize the outputs
Solution Approach 1:
The patent performs aggregation operations in advance and stores the results in a horizontal format that is pre-suited for analytical applications. By pre-computing and pre-formatting the aggregation results horizontally during the database query execution, the system eliminates the need for subsequent manual or automated summarization processes, thereby saving significant time and effort in data analysis workflows.
3Productivity
If conventional group-by operations are performed on very large relational tables, then aggregation results are generated, but the output produces a large number of rows that are unwieldy
Solution Approach 1:
The patent merges multiple aggregate values into the same output row by arranging them in a horizontal format. Instead of producing separate rows for each aggregate value, the system combines multiple aggregation results (such as sum, average, count, min, max for different grouping columns) into a single horizontal record, thereby reducing the total number of output rows and simplifying the overall output structure while maintaining computational efficiency.
Data Source
AI summary
A database system is able to receive a query containing a horizontal aggregate function. In response to the query containing the horizontal aggregate function, aggregate values are produced in a horizontal format. In general, the query can contain both horizontal (new) aggregate functions and vertical (standard) aggregate functions.