Parallel Coordinates Visualization Using Bi-variant Bins
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Handling large amounts of multivariate data in parallel coordinates visualization is cumbersome due to the complexity of plotting numerous records, leading to indecipherable graphs with crowded plots of lines, which hinders effective data analysis.
Innovation Solution
The method involves segmenting parallel coordinates into buckets, aggregating records from a database, and plotting them on a graph to generate a visualization, using bi-variant bins and statistical analysis to infer data flow and relationships between categorical and continuous variables.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If every record is fetched from data store and plotted as a graph of parallel co-ordinates, then complete data visualization is achieved, but handling enormous amount of data becomes cumbersome and the graph becomes indecipherable
Solution Approach 1:
The patent segments the parallel coordinates graph into multiple sub-graphs by dividing the data range into buckets. Each sub-graph handles a portion of the data, making the visualization manageable and decipherable while maintaining complete data representation through the collection of all sub-graphs.
Solution Approach 2:
The patent extracts and plots only the necessary data points that fall within specific bucket ranges rather than plotting all records. This selective extraction reduces the number of lines in each sub-graph, making the visualization easier to interpret while preserving the essential data relationships.
2Loss of information
If every record is plotted as a line in parallel co-ordinates, then complete data representation is achieved, but the graph becomes crowded and indecipherable
Solution Approach 1:
The patent divides the parallel coordinates graph into multiple sub-graphs based on bucket ranges. Each sub-graph contains fewer lines representing data points within its specific range, reducing the visual complexity and crowding while maintaining complete data representation across all sub-graphs.
Solution Approach 2:
The patent introduces a dimensional separation by creating multiple sub-graphs arranged in a structured layout. This additional organizational dimension allows the viewer to process information systematically across sub-graphs rather than being overwhelmed by a single dense graph.
3Ease of operation
If the graph is segmented into buckets with aggregation, then data handling ease is improved, but the level of detail in individual records is reduced
Solution Approach 1:
The patent segments data into buckets with specific range definitions, allowing efficient data handling and processing. Each bucket maintains its own sub-graph with aggregated data points, preserving the ability to analyze individual records within the context of the overall data distribution.
Solution Approach 2:
The patent performs preliminary aggregation of data points within each bucket before plotting. This preliminary action organizes the data in advance, making the visualization easier to handle while maintaining the ability to access and analyze individual record details when needed through the structured bucket organization.
Data Source
AI summary
Parallel co-ordinates associated with a business scenario represent underlying data of the business scenario. Based upon associated categorical variables, the parallel co-ordinates are divided into a plurality of segments. Each segment includes a pair of continuous variables represented as a pair of axes of the parallel co-ordinates. The pairs of axes of the parallel co-ordinates are fragmented to generate bi-variant bins. The bi-variant bines include continuous values corresponding to each fragment of the pair of axis. An associated backend database is queried to determine aggregated tuples of continuous values between each bi-variant bin of the corresponding pair of axes. Based on the aggregated tuples, a statistical visualization of the aggregated tuples between the bi-variant bins for the axes is generating. The statistical visualization includes a distribution of the tuples for the plurality of segments corresponding to the parallel co-ordinates.