Time Series Visualization via M4 Aggregation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large volumes of time series data from sources like sensor networks and financial markets pose challenges in visualization, as existing methods often result in high bandwidth consumption and processing loads due to the need for complex queries and detailed data representation, which can impair visualization efficiency.
Innovation Solution
The method involves preprocessing time series data to generate tuples with minimum, maximum, first, and last values for each time interval, allowing for efficient data reduction and visualization by selecting the appropriate time durations and applying M4 aggregation, thereby optimizing data access and visualization within a browser window.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If detailed time series data is transmitted from data sources to the browser for visualization, then visualization accuracy is improved, but bandwidth consumption increases
Solution Approach 1:
The patent extracts only the essential characteristics of time series data (minimum, maximum, first, and last values within time bins) rather than transmitting complete detailed data. This extraction approach maintains visualization accuracy by preserving key data features while significantly reducing bandwidth consumption by filtering out redundant information before transmission to the browser.
Solution Approach 2:
The patent performs preliminary aggregation of time series data at the data source before transmission. By pre-computing the four-tuple representations (minimum, maximum, first, last values) and organizing data into time bins, the system prepares optimized data structures in advance, reducing the amount of data that needs to be transmitted and processed during visualization while maintaining accuracy.
2Measurement precision
If complete time series data is processed in the browser for visualization, then data accuracy is improved, but processing load increases
Solution Approach 1:
The patent extracts only the essential characteristics of time series data (minimum, maximum, first, and last values within time bins) rather than transmitting complete detailed data. This extraction approach maintains visualization accuracy by preserving key data features while significantly reducing bandwidth consumption by filtering out redundant information before transmission to the browser.
Solution Approach 2:
The patent performs preliminary aggregation of time series data at the data source before transmission. By pre-computing the four-tuple representations (minimum, maximum, first, last values) and organizing data into time bins, the system prepares optimized data structures in advance, reducing the amount of data that needs to be transmitted and processed during visualization while maintaining accuracy.
3Quantity of substance
If complex queries are issued to retrieve time series data for visualization, then data completeness is improved, but data access complexity increases
Solution Approach 1:
The patent performs preliminary aggregation of time series data at the data source before transmission. By pre-computing the four-tuple representations (minimum, maximum, first, last values) and organizing data into time bins, the system prepares optimized data structures in advance, reducing the amount of data that needs to be transmitted and processed during visualization while maintaining accuracy.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure relates to a method comprising: • sending to the data source a request for visualizing sensor data within a canvas having a width of w pixels and covering a visualization time range, each pixel of the canvas being representative of a time duration Iopt, • receiving consecutive sets of tuples that each covers a time interval having the time duration It, • performing a M4 aggregation comprising - generating, from the received tuples, a set of consecutive w groups, each group of the w groups comprising tuples covering a time interval having the time duration Iopt, - determining for each group of the w groups a set of aggregates, • displaying the w sets of aggregates on the canvas of the browser window as a chart, wherein each one of the sets of aggregates is displayed in one of pixel columns of the canvas.