Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Advanced DataVisualization and Analysis:
WebFOCUS Visual Discovery
A Technology Brief by Jodi Keller
Data Visualization Overview
Types of Visualization and Their Uses
Bar Charts
Line Charts
Pie Charts
Counts
Data Constellations
Data Constellation Maps
Data Sheets
Histograms
Multiscapes
Paraboxes
Scatterplots
Time Tables
Perspectives
WebFOCUS Visual Discovery in Action: An Example
Conclusion
1
2
2
14
18
3
4
5
6
7
8
8
9
10
11
12
12
Table of Contents
Edward Tufte, a Yale professor and a pioneer in the concepts of data illustration and visualization,
posed in his book, Envisioning Information, the following problem: the world is dynamic and
multidimensional, but both paper and computer screens are static and flat. How do you represent
such a rich world of experience and measurement on a mere flatland?
This challenge also applies to today’s businesses, where operations are complex and multifaceted.
How can the large volume of data created by intricate business transactions and activities be
conveyed in a way that can be easily understood and analyzed? Will simple tables and reports be
enough, or will they limit the business users’ ability to interpret the context of that information and
understand its meaning and impact – rendering it virtually useless?
Data visualization turns large quantities of raw data into perceptive images, that provide a
powerful and dynamic instrument for the discovery and reasoning of quantitative information. It
significantly increases the interpretability and usability of complex business data by displaying it as
geometric shapes (points, lines, circles, rectangles, etc.) with various attributes (proximity, size,
color, etc.) that represent its value or meaning. And, when information is displayed as sophisticated
graphical illustrations, and therefore communicated in a clearer and more precise manner, it
instantly becomes more valuable. As a result, users can do more than just read critical business
information – they can envision it, make sense of it, and leverage it to gain a competitive edge.
WebFOCUS Visual Discovery is a robust solution that provides powerful, real-time visualization
capabilities. Users at all levels can easily perform intuitive, highly interactive, graphical analysis
from any enterprise data source using a wide variety of advanced visualization mechanisms and
components, including histograms, data constellations, and multiscapes. With WebFOCUS Visual
Discovery, data can be more easily measured and compared, critical trends and relationships can
be spotted in an instant, and problems can be immediately detected and acted upon.
Data Visualization Overview
1 Information Builders
2 WebFOCUS Visual Discovery White Paper
Types of Visualization and Their UsesTables, charts, and graphs communicate information in an illustrative way and help make
complex business data easy to read and understand. But in order for these images to be as
intuitive as possible, the information they contain must be displayed in the most appropriate
format. Otherwise, they can cause confusion and create what Tufte refers to as “statistical
murk.” On the other hand, the correct graphical representation of business data – whether it’s
pie charts, scatterplots, or histograms – can give new meaning to the information users view
and can provide instantaneous answers to the most urgent business questions.
Bar Charts
A bar chart, or bar graph, is a diagram that provides a means of quantitative comparison by
displaying rectangular bars of differing heights or lengths that are proportional to the statistics or
data they represent. Bar charts display the distribution of a single, discrete or dependent variable
across an independent variable, and can be arranged either horizontally or vertically. (These are
also known as column charts or graphs.) Users most frequently use bar charts to analyze changes
within a data group or category or to demonstrate dissimilarities between items.
With WebFOCUS Visual Discovery, users can measure and compare information by creating
comprehensive, easily understandable bar charts. These bar charts can support thousands of
bars and can include robust features such as:
■ Stacked colors – simultaneously display multiple attributes in a single bar or column
You would use a barchart to measure:■ Total expenses by
month
■ Total number ofemployees by salarygroup
■ Sales or revenue byproduct line
3 Information Builders
■ Sorting – quickly detect patterns and find results
■ Zooming – select and view appropriate subsets of data
■ Spine-plot variation – distinguish proportions across groups by simply adjusting bar
thickness based on the data
■ Goal lines – accurately measure actual versus planned results
■ Weighting – display data values in accordance to bias
■ Animation control
Line Charts
A line chart, or line graph, is a diagram that compares the relationship between two variables –
one dependent and one independent. It demonstrates the effect of the independent variable on
the dependent variable by showing successive points, each representing a different data value,
connected by straight lines. Line charts are particularly useful in examining a sequence of values
to determine growth trends and displaying data that changes continuously over time. They can
also help users identify patterns that make predictions about the results of data not yet recorded.
WebFOCUS Visual Discovery enables users to rapidly create a variety of line charts, including
filled line charts, as well as multiple line charts that are directly stacked, or stacked in separate
You would use a linechart to measure:■ Revenue by quarter
■ Total number ofproduct returns bymonth
■ Incoming calls to callcenter by hour
4 WebFOCUS Visual Discovery White Paper
panels. Additionally, users can display multiple lines within the same line chart and scroll or
zoom for convenient navigation of long lines. Other capabilities include:
■ Stacked colors – simultaneously display multiple attributes in a single line
■ Clusters – group related lines together on the diagram
■ Weighting – assess data values on a “curve”
■ Animation control
Pie Charts
Pie charts, also known as pie graphs, circular charts, or circular graphs, divide complete data
sets into “slices,” with each slice representing a group or subset of that data. The size of the
slice corresponds with the percentage of the whole that particular group makes up. Pie charts
are best for understanding data in terms of proportions and showing how sizes of parts relate to
a whole.
You would use a piechart to see:■ What percentage of
total sales eachproduct linerepresents
■ What percentage ofthe total revenuescomes from each sales region
■ How much of a totalbudget is allocated toparticular items
5 Information Builders
WebFOCUS Visual Discovery enables users to create powerful pie charts or view groups of pie
charts to compare a number of categories at once. Users can also customize their view of the
pie chart through flexible color-coding, rotation, exploding, and ordering. Additional features
include:
■ Stacked colors – simultaneously display multiple attributes in a single slice
■ Weighting – assess data values on a bias
Counts
A count provides a statistical summary of a data table, and lists exact numerical values for data
fields within the table, in an easy-to-read, textual format. Counts should be used when specific
information (e.g., mean, standard deviation, minimum, maximum, or most common value) is
needed for either an entire table or a selected subset of that table.
Counts that are created using WebFOCUS Visual Discovery can also include:
■ Selected or deleted records
■ Total records
■ Miscellaneous statistics
Data Constellations
Data constellations, or constellation graphs, depict multiple data sets – or groups of data sets
with particular attributes – and map the association between them. They are called
“constellations” because they are made up of numerous “stars” or nodes – each representing a
specific data value – that are distanced from and connected to each other on the diagram plane
in proportion to the strength or weakness of their relationship. Data constellations are best
suited for plotting and analyzing large, complex data sets in a single view. Additional data
attributes and relationships can also be displayed by adjusting node and link size, style, and
color based on data values.
With WebFOCUS’ visualization features, users can fully understand and explore relationships by
creating complete, easy-to-read data constellations. Users can further enhance these
constellations by:
■ Controlling node and link appearance based on data values
■ Positioning nodes automatically, based on the strength of the relationship, or manually,
based on preferences or other criteria
■ Labeling nodes and links for easier identification
6 WebFOCUS Visual Discovery White Paper
You would use a dataconstellation to analyze:■ Purchasing
associations
■ Web click streams
■ Viral-marketing-campaign patterns
7 Information Builders
Data Constellation Maps
When analysis of information with a geographical or spatial component is required, a data
constellation can also be presented as a map by associating the nodes on the constellation with
map images and geo-code data such as countries, cities, zip codes, and latitude and longitude
coordinates. Data constellations maps are helpful when users need to distinguish regional
variations in specific attributes or determine links between specific locations. As with standard
data constellations, the height, shape, size, and color of nodes and links on a constellation map
can be manipulated to represent other data relationships or characteristics.
WebFOCUS Visual Discovery includes complete maps of the world, as well as the United States,
Canada and Mexico, with geo-code data for countries, over 400 worldwide cities, and U.S. zip
codes. Custom maps, such as floor plans and other spatial images, can also be supported by
using bitmap images as a base with associated geo-code data. Users can also:
■ Control node and link appearance based on data values
■ Position nodes based on the map image
■ Color-code specific regions on the map or globe
■ Label nodes and links for easier identification
You can use a dataconstellation map to view:■ Customer population
by state
■ A “visual supplychain” (e.g., most frequently usedtransportation routesbetween keydistribution centers)
8 WebFOCUS Visual Discovery White Paper
Data Sheets
A data sheet creates a table-style view of a set of fields so users can easily uncover interactions
and patterns. Data sheets are particularly useful for viewing dynamic “top” or “bottom” lists or
dividing data into groups for comparison to identify both common and unique attributes.
Data sheets created using WebFOCUS Visual Discovery allow users to:
■ Pan and zoom to alter viewpoint
■ “Squash” the data to reveal patterns across several variables by comparing hundreds to
thousands of rows at once
■ Interact to identify interesting sub-network
■ Combine rows into line representation
■ Sort the table by one or more fields
■ Perform textual searches using specific text or substrings of text
■ Export selected subsets of rows and columns of data
Histograms
A histogram groups data values into classes and shows the frequency at which each class
appears in the data set by displaying a series of columns whose width represents class intervals
and whose areas are proportional to the corresponding frequencies. The shapes of histograms
You can use data sheets to:■ See your best-selling
products
■ Know who your topperforming sales reps are
■ Compare customercharacteristics based on products purchased
9 Information Builders
will vary depending on the choice of the size of the intervals. Histograms are designed for
situations when the user is not interested in knowing the particular values, but in how those
values are distributed across the data set.
WebFOCUS Visual Discovery empowers users to see and understand locations and variations
within their data by creating easy-to-read histograms with powerful features such as:
■ Stacked colors – simultaneously represent multiple attributes of a single data class or group
■ Smoothing – control the way random variations in the data are displayed
■ Orientation – display the histogram either vertically or horizontally
■ Weighting – display class or group values in accordance to a bias
Multiscapes
Multiscapes provide a visual representation of the interaction or relationship between two or
more associated data fields by plotting one variable on the x-axis and one or more variables on
the y-axis. Multiscapes are a very effective means for understanding large bar charts with
multiple categories, and are typically used to compare objects based on several attributes at
the same time.
A histogram can helpyou analyze:■ Number of employees
by salary range
■ Number of sales byprice range
■ Number of calls to acall center by timeperiod
10 WebFOCUS Visual Discovery White Paper
With WebFOCUS Visual Discovery, users can generate two- or three-dimensional multiscapes to
rapidly find patterns between sets of statistics characterizing an object. These multiscapes can
accommodate scenarios with hundreds of bars along each axis, and users can touch any
individual bar to highlight its categories, sweep across a range of bars to see detail, or
interactively isolate groups of bars by choosing other attributes. Other features include:
■ Multiple bar shapes
■ Ordering and clustering – group similar data values together on the diagram
■ Panning, zooming, and rotating – view information from multiple points of view and
conveniently navigate through large three-dimensional diagrams
■ Weighting – display bars in accordance to a bias
Paraboxes
A parabox is a hybrid of box plots and parallel coordinate plots, and is made up of multiple
columns – the first containing specific values from a data group (each displayed as a bubble,
sized to demonstrate the number of records it represents), with the remaining columns each
representing a specific characteristic of that group. Paraboxes can give users an overview of the
distribution of values in a set of fields from a table and provide a simple summary of how a
subset of data differs from the whole. Paraboxes are effective for simultaneously measuring and
analyzing numerous attributes and complex relationships among multiple fields.
Multiscapes can be used to measure:■ Customers by region,
industry, and compa-ny size
■ Revenues by country,sales channel, andproduct
■ Employees by age,race, and length ofservice
11 Information Builders
WebFOCUS Visual Discovery offers a unique parabox component. Users can create intuitive
paraboxes, then further enhance them by drawing lines that connect column values, so
interaction patterns can be detected between fields and abnormalities that fall outside the regular
range can be identified. Users can also perform comparisons by population or selection, access
individual data values, and choose from multiple plot styles.
Scatterplots
A scatterplot, also called a scattergram or scatter diagram, is used to investigate the possible
association between two variables that both relate to the same event by distributing a series of
points or nodes, each representing a value in the data set, across a grid. Scatterplots are used
to demonstrate cause and effect and evaluate the nature and degree of associations between
two attributes.
With WebFOCUS Visual Discovery users can easily visualize correlations between two items.
Additionally, scatterplots can be enhanced by:
You would use a paraboxto see:■ Price, cost of goods
sold, profit margin,YTD sales, and growthpercentage overprevious year for eachproduct within aspecific product line
■ Revenue generated,YTD number of deals,YTD number of salesto new-name accounts,and total number ofcustomers for eachsales rep within aparticular sales region
Scatterplots can helpyou analyze:■ The frequency of
consumer purchasescorrelated with thereach of advertising
■ The associationbetween employees’level of education andincome
■ The relationshipbetween revenue by customer and customercompany size
12 WebFOCUS Visual Discovery White Paper
■ Adjusting node size, shape, and color
■ Superimposing trend lines to highlight patterns in the whole data set or selected items
■ Including jitter nodes to show overplotting
■ Adding nonlinear axes
Time Tables
Time tables show activities or tasks over a period of time. Tick marks representing individual
data records are displayed across a grid in a pattern directly related to the time of their
occurrence. Time tables are used to show large numbers of events over time and for studying
the interrelationships between those events.
WebFOCUS Visual Discovery’s time table component allows users to monitor the timing and
pattern of vital business activities. Users can use shaped or angled glyphs or symbols to
represent specific activities, group related ticks together on the grid, and even adjust the size
and color of ticks to indicate event duration.
Perspectives
Users can link a combination of visualization components and place them together on a single
page or screen to obtain a simultaneous view of multiple facets of critical information.
Perspectives can be created on-the-fly, to answer immediate business questions, or reused for
other reporting and monitoring tasks.
With time tables, youcan quickly see a varietyof patterns:■ Incoming calls to a
call center
■ Hits on a Web site
■ Responses tosuccessive marketingcampaigns
14 WebFOCUS Visual Discovery White Paper
WebFOCUS Visual Discovery In Action: An Example
WebFOCUS Visual Discovery can be used to make many interactive discoveries from a single,
portal-type interface. In this example, we are analysts at Century Corp., a consumer electronics
company, who are trying to better understand purchasing trends. We begin our analysis by
viewing sales information related to the various products we sell.
In the upper right-hand corner, a scatterplot shows a clustering of dots representing orders for
product, which can be used to easily detect patterns. Order revenue (in thousands) is
distributed on the X-axis, and the margin obtained from those orders is shown on the y-axis. Dot
color varies to demonstrate the relationship between the orders and the stores at which those
orders were taken.
15 Information Builders
By selecting an area on the scatterplot, we can focus on certain information contained in a
specific dot cluster. For example, the area selected above represents orders with a small dollar
value (between $0 and $60,000) but high margin (greater than 20 percent).
The area we highlight on the scatterplot will also affect the other charts on the screen, so those
graphs will change to show only the selected data set. In the bar chart (top, left-hand corner),
we can see that the orders we are interested in are coming primarily from the Web, where
revenue is typically minimal, but overhead is relatively low.
The data sheet in the bottom left corner displays other metrics regarding the selected orders,
specifically the revenue and margin for each separate order number within that group.
16 WebFOCUS Visual Discovery White Paper
The drop-down list at the top of our console enables us to filter the products we’re looking at.
For example, we can alter our graphs and charts to provide data related only to DVDs, instead of
all products sold by Century Corp. By making that change, we can now see which retailers are
selling the highest percentage of DVDs in the bar chart, DVD ordering patterns in the scatter-
plot, and revenue and margins for each separate DVD order number in the data sheet. By
analyzing the scatterplot, we realize that the majority of DVD sales have a low dollar value.
17 Information Builders
Now, let’s focus on the multiscape in the bottom right corner of our console. We can swivel the
multiscape to see the revenue for each store by year and quarter. This enables us to recognize
that sales at eMart have been on the rise, while sales at other retailers have dipped in recent
quarters.
We’ve now seen how the usage and interaction of these combined data visualization
components can provide hundreds of observations about data relationships – relationships that
would be virtually impossible to see if the data were represented in simple reports and tables.
18 WebFOCUS Visual Discovery White Paper
ConclusionData visualization turns raw data info intuitive images, significantly increasing a user’s ability to
measure, compare, observe, and analyze complex business information. But in order for data
visualization to be truly effective, users must understand the various techniques available, as
well as their appropriate uses and ensure they are illustrating information in the most suitable
graphical format.
WebFOCUS Visual Discovery makes it easy for users to visualize and analyze their most critical
enterprise data in an almost unlimited number of ways – from simple line and bar charts to
more complex illustrations such as scatterplots and paraboxes. With WebFOCUS Visual
Discovery, vital information becomes more meaningful, understandable, and usable than ever
before, so critical business questions can be answered immediately, patterns and relationships
can be rapidly identified, and problems can be instantly detected and corrected.
North AmericaUnited States■ Atlanta,* GA (770) 395-9913
■ Baltimore, MD Consulting: (703) 247-5565
■ Boston,* MA (781) 224-7660
■ Charlotte,* NC Consulting: (704) 494-2680
■ Chicago,* IL (630) 971-6700
■ Cincinnati,* OH (513) 891-2338
■ Cleveland,* OH (216) 520-1333
■ Dallas,* TX (972) 490-1300
■ Denver,* CO (303) 770-4440
■ Detroit,* MI (248) 743-3030
■ Federal Systems,* DC (703) 276-9006
■ Hartford, CT (860) 249-7229
■ Houston,* TX (713) 952-4800
■ Los Angeles,* CA (310) 615-0735
■ Metropolitan,* NY Sales: (212) 736-7928
Consulting: (212) 736-4433, ext. 4443
■ Minneapolis,* MN (651) 602-9100
■ New Jersey* (973) 593-0022
■ Orlando,* FL (407) 804-8000
■ Philadelphia,* PA (610) 940-0790
■ Pittsburgh, PA (412) 494-9699
■ St. Louis,* MO (636) 519-1411
■ San Jose,* CA (408) 453-7600
■ Seattle,* WA (206) 624-9055
■ Washington,* DC Sales: (703) 276-9006
Consulting: (703) 247-5565
CanadaInformation Builders (Canada) Inc.
■ Calgary (403) 538-5415
■ Montreal* (514) 630-1134
■ Ottawa (613) 233-0865
■ Toronto* (416) 364-2760
■ Vancouver* (604) 688-2499
■ Victoria (250) 995-8674
MexicoInformation Builders Mexico
■ Mexico City 52-55-91-71-14-61
Europe■ Belgium Information Builders Belgium
Brussels 32-2-7430240
■ France Information Builders France S.A.
Paris 33-14-507-6600
■ Germany Information Builders (Deutschland)
Dusseldorf 49-211-522877-0
Eschborn 49-6196-77576-0
Munich 49-89-35489-0
Stuttgart 49-711-7287288-0
■ Netherlands Information Builders
(Netherlands) B.V.
Amsterdam 31-20-4563333
■ Portugal Information Builders Portugal
Lisbon 351-217-230-720
■ Spain Information Builders Iberica S.A.
Barcelona 34-93-344-32-70
Bilbao 34-94-425-72-24
Madrid 34-91-710-22-75
■ Switzerland Information Builders
Switzerland AG
Wallisellen 41-1-8394949
■ United Kingdom Information Builders
(UK) Ltd.
London 44-208-9824700
Warrington 44-1925-820111
AustraliaInformation Builders Pty. Ltd.
■ Melbourne 61-3-9631-7900
■ Sydney 61-2-8223-0600
Representatives■ Austria FOCUS Informationstechnologie GmbH
Vienna 43-12-1136-3870■ Brazil InfoBuild Brazil
São Paulo 55-11-3017-5178■ China InfoBuild China, Inc.
Shanghai 86-21-5080-5431
■ Colombia Coransa
Bogota 571-635-0273
■ Finland InfoBuild Oy
Helsinki 358-9-7250-2250
■ Greece Applied Science
Athens 30-210-699-8225■ Guatemala IDS de Centroamerica
Guatemala City 502-361-0506
■ Gulf States ■ Bahrain ■ Kuwait ■ Oman
■ Qatar ■ Saudi Arabia ■ Yemen
■ United Arab Emirates
Al-Gosaibi Information Systems
973-274-090
■ Israel NESS A.T. Ltd.
Tel Aviv 972-3-5483638
■ Italy Selesta G C Applications S.P.A.
Genova 39-010-64201-224
Milan 39-02-2515181
Torino 39-011-5513-211■ Japan K.K. Ashisuto
Osaka 81-6-6373-7113
Tokyo 81-3-3437-0651
Simplex Solutions
Tokyo 81-3-5652-3180/81
■ Korea Unitech Infocom Co. Ltd.
Seoul 82-2-3477-4456■ Malaysia Optegra Sdn Bhd
Selangor 60-3-80240188
■ Norway iSolutions AS
Stavanger 47-51-44-47-44
■ Philippines
Beacon Frontline Solutions, Inc.
63-2-750-1972
Corporate Information solution
63-2-633-1321
■ Poland Compfort/Meridian Polska SP
Warsaw 4822-630-2660
■ Singapore
Automatic Identification Technology Ltd.
65-6286-2922
Legato Solutions and Services Pte Ltd.
65-684-63150
■ South Africa International Computers S.A.
(Pty.) Ltd.
Johannesburg 27-11-2335911
■ Sweden Cybernetics Business Solutions AB
Solna 46-7539900
■ Taiwan Galaxy Software Services
Taipei 886-22-3897722
■ Turkey Istanbul
Erdemsoft 90-212-257-5555
Key Soft Ltd. 90-216-428-5933
■ Venezuela InfoServices Consulting
Caracas 58-212-763-1653
Toll-Free Numbers■ Sales and Information
(800) 969-INFO
■ VAR and Reseller Information
(800) 969-4636
*Training facilities are located at these branches;additional locations are available.
Corporate Headquarters Two Penn Plaza, New York, NY 10121-2898 (212) 736-4433 Fax (212) 967-6406 DN7504242.1203
www.informationbuilders.com [email protected] Headquarters 150 York St., Suite 1000, Toronto, ON M5H 3S5 (416) 364-2760 Fax (416) 364-6552 For International Inquiries +1(212) 736-4433
Copyright © 2003 by Information Builders, Inc. All rights reserved. [26]All products and product names mentioned in this publication are trademarks or registered trademarks of their respective companies. Printed in the U.S.A.
on recycled paper
Sales and Consulting Offices