Hybrid Chart Segmentation for Big Data Visualization

Solution Overview

Problem

Traditional two-dimensional chart methods, such as scatterplots and histograms, are inadequate for large datasets as they become computationally expensive and cluttered, making it difficult to extract meaningful information, and hinder critical analysis by preventing individual object examination.

Innovation Solution

A hybrid chart is generated by dividing the dataset into segments, where some segments display data like a scatterplot and others like a histogram, allowing dynamic switching between chart types, reducing processing power and memory requirements, and enabling user interaction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a scatterplot is used to display each object in the dataset, then individual objects can be examined, but the chart becomes cluttered and computationally expensive with large datasets

Engineering Contradiction:

Improveindividual object examinationVSAvoidchart clutter and computational cost

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The chart is divided into multiple segments or regions, each capable of displaying different numbers of data points. High-density regions use aggregated displays while low-density regions show individual points, allowing the system to manage complexity while preserving individual object examination where needed.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Different portions of the chart apply different display strategies based on local data density. Some segments display individual points while others use aggregated representations, optimizing both individual object examination and overall chart manageability by adapting to local conditions.

Inventive Principle:

Principle #3Local quality

2Device complexity

If a histogram is used to group objects into segments, then the chart is less cluttered, but critical analysis is prevented since objects cannot be individually examined

Engineering Contradiction:

Improvechart clutter reductionVSAvoidindividual object examination

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The chart is divided into multiple segments or regions, each capable of displaying different numbers of data points. High-density regions use aggregated displays while low-density regions show individual points, allowing the system to manage complexity while preserving individual object examination where needed.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The display mode of each segment is dynamic and can switch between aggregated histogram-style display and individual point display based on user interaction, data density, or zoom level. This allows the chart to maintain low clutter while enabling individual object examination when required.

Inventive Principle:

Principle #15Dynamics

3Loss of information

If all data points are plotted individually, then complete data visibility is achieved, but processing power and memory requirements increase extensively

Engineering Contradiction:

Improvedata visibilityVSAvoidprocessing power and memory usage

Core Design Contradiction:

Loss of informationVSUse of energy by moving object

Solution Approach 1:

The chart is divided into multiple segments or regions, each capable of displaying different numbers of data points. High-density regions use aggregated displays while low-density regions show individual points, allowing the system to manage complexity while preserving individual object examination where needed.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Instead of plotting all data points individually across the entire chart, the system applies individual point plotting only to segments where it provides value (low-density regions), while using aggregated display for high-density regions. This partial application of the detailed display strategy reduces processing and memory requirements while maintaining data visibility where most needed.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS10169889B2Techniques for visualizing big data in a two-dimensional chart

Publication Date: 2019.01.01 SAP SE

AI summary

A technique is described for analyzing a dataset that includes many dataset objects by generating a hybrid chart. The hybrid chart is a two-dimensional chart that is divided up into multiple segments. Some segments are configured to present a color from a color scale that is used to represent the number of dataset objects that lie within the segment. Other segments are configured to present dataset objects within the segment by plotting the dataset objects. Depending on the number of dataset objects within the segment, one of these two approaches can be used. The chart offers interactivity like zooming and selecting a range of datasets. Furthermore the chart starts by offering a coarse overview while generating views with increasing granularity and more details over time. During this process the chart remains completely interactive.