Browse Prior Art Database

Word-sized Graphic Representation of Data Distributions

IP.com Disclosure Number: IPCOM000218197D
Publication Date: 2012-May-26
Document File: 4 page(s) / 124K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method to provide a pattern visualization capability in a chart/graphic that fits within a table or dashboard display.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 4

Word-sized Graphic Representation of Data Distributions

Many user interfaces (UI) have a need to show complex data with a small screen real estate; dashboards are a classic example. It is often desired to show not only a simple number, but also some historical chart to show trending information. Typically, this is solved using a sparkline approach. (Figure 1)

Figure 1: Sparkline approach to illustrate trending

This works very well for time-series data (e.g., stock prices, temperature, etc.); however, it does not display data distributions such as shown in an x-y scatter diagram. (Figure


2)

Figure 2: x-y scatter diagram

1


Page 02 of 4

This diagram shows a set of assets (e.g., computer systems, say) over time by some metric (e.g., percentage of utilization). A dashed threshold line indicates a point where the samples cross into a "danger zone" and require special attention. The typical spark-line approach does not work in this case because it is not simply a line chart, but a series of data points scattered across the domain that really represents the data distribution. The x-y scatter diagram is showing the distribution of data, the clusters of samples, and how those clusters shift through time. It is not just a simple value over time.

When a user looks at Figure 2, they can visualize a pattern where there are two clusters (one cluster starts down near the bottom of the y-axis, and the other cluster starting around the mid-way point) converging together over time.

A method is needed to provide a similar capability in a chart/graphic that fits within a table or dashboard display.

The idea is to divide the time-axis into time ranges or equal size (i.e., time-buckets) and represent one bucket at a time, shown horizontally instead of vertically. With animation, the user can cycle forward (or backward) through the time-buckets to show how the distribution shifts over time. This slicing of the 2-dimensional space allows the graphic display to fit into a word-sized element suitable for dashboards and other information-dense displays.

There are a number of ways to represent it. Perhaps the simplest mechanism is to take that time-bucket slice through the x-y chart and represent it as a horizontal 1-dimensiona...