Information systems final - chunk 2

extract

gather data from multiple sources and pull it into a staging area for processing

transform

clean, reformat, and standardize the data, fixing errors and applying business rules so it is convenient and analysis ready

load

move the prepared data into its destination system for storage, querying, and reporting

4 types of data visualizations

idea illustration, everyday dataviz, idea generation, visual discovery

graph descriptor for visualizations

conceptual, data driven, declarative, exploratory

conceptual

focused on ideas

data driven

focused on statistics

declarative

focused on documenting and designing

exploratory

focused on prototyping, iterating, interacting, and automating

idea illustration goals

learning, simplifying, explaining

idea illustration description

“consultant’s corner,” used to clarify complex ideas by drawing on our ability to understand metaphors and simple design conventions

idea illustration on map

conceptual and declarative

idea generation goals

problem solving, discovery, and innovation

idea generation on map

conceptual and exploratory

idea generation description

used to find new ways of seeing how a business works and to answer complex managerial challenges

visual discovery goals

trend spotting, sense making, and deep analysis

visual discovery on map

data driven and exploratory (big data and complex)

visual discovery description

lends itself to interactivity (injecting new data sources to continually revisualize), often produces insights that can’t easily be discovered by looking at the raw data, function trumps form

everyday dataviz goals

affirming and setting context

everyday dataviz on map

data driven and declarative

everyday dataviz description

used for formal, storytelling presentations, usually simple and communicate a single message, should speak for itself

importance of effective visualizations

1. increase processing speed

2. reduce time to insight

3. some data makes more sense

reducing time to insight

allows clear conclusions to be understood, speaks for itself, saves time

best data visualization practices

1. tell a story

2. maintain graphic integrity

3. minimize graphical complexity

maintaining graphical integrity

1. know when text is best

2. include a title

3. start axes at 0

4. make sure circles are sized appropriately

minimizing complexity

avoid chartjunk

best practices for communicating data insights

1. state your key point

2. be complete, yet concise

3. avoid unnecessary clutter

4. acknowledge data source

color guidelines

1. color choices should make sense (red = hot)

2. limit amount of colors for discrete data

3. consider the color blind (orange and blue diverging is better than red and green)